Add RuntimeManager to centralize runtime management

.dockerignore

@@ -1,5 +0,0 @@
-frontend/node_modules
-config.toml
-.envrc
-.env
-.git

.gitattributes

@@ -1 +0,0 @@
-*.ipynb linguist-vendored

.github/.codecov.yml

@@ -1,19 +0,0 @@
-codecov:
-  notify:
-    wait_for_ci: true
-    # our project is large, so 6 builds are typically uploaded. this waits till 5/6
-    # See https://docs.codecov.com/docs/notifications#section-preventing-notifications-until-after-n-builds
-    after_n_builds: 5
-
-coverage:
-  status:
-    patch:
-      default:
-        threshold: 100% # allow patch coverage to be lower than project coverage by any amount
-    project:
-      default:
-        threshold: 5% # allow project coverage to drop at most 5%
-
-comment: false
-github_checks:
-    annotations: false

.github/ISSUE_TEMPLATE/bug_template.yml

@@ -1,62 +0,0 @@
-name: Bug
-description: Report a problem with OpenHands
-title: '[Bug]: '
-labels: ['bug']
-body:
-  - type: markdown
-    attributes:
-      value: Thank you for taking the time to fill out this bug report. Please provide as much information as possible to help us understand and address the issue effectively.
-
-  - type: checkboxes
-    attributes:
-      label: Is there an existing issue for the same bug?
-      description: Please check if an issue already exists for the bug you encountered.
-      options:
-      - label: I have checked the existing issues.
-        required: true
-
-  - type: textarea
-    id: bug-description
-    attributes:
-      label: Describe the bug and reproduction steps
-      description: Provide a description of the issue along with any reproduction steps.
-    validations:
-      required: true
-
-  - type: dropdown
-    id: installation
-    attributes:
-      label: OpenHands Installation
-      description: How are you running OpenHands?
-      options:
-        - Docker command in README
-        - GitHub resolver
-        - Development workflow
-        - app.all-hands.dev
-        - Other
-      default: 0
-
-  - type: input
-    id: openhands-version
-    attributes:
-      label: OpenHands Version
-      description: What version of OpenHands are you using?
-      placeholder: ex. 0.9.8, main, etc.
-
-  - type: dropdown
-    id: os
-    attributes:
-      label: Operating System
-      options:
-        - MacOS
-        - Linux
-        - WSL on Windows
-
-  - type: textarea
-    id: additional-context
-    attributes:
-      label: Logs, Errors, Screenshots, and Additional Context
-      description: Please provide any additional information you think might help. If you want to share the chat history
-        you can click the thumbs-down (👎) button above the input field and you will get a shareable link
-        (you can also click thumbs up when things are going well of course!). LLM logs will be stored in the
-        `logs/llm/default` folder. Please add any additional context about the problem here.

.github/ISSUE_TEMPLATE/feature_request.md

@@ -1,18 +0,0 @@
----
-name: Feature Request
-about: Suggest an idea for OpenHands features
-title: ''
-labels: 'enhancement'
-assignees: ''
-
----
-
-**What problem or use case are you trying to solve?**
-
-**Describe the UX of the solution you'd like**
-
-**Do you have thoughts on the technical implementation?**
-
-**Describe alternatives you've considered**
-
-**Additional context**

.github/ISSUE_TEMPLATE/technical_proposal.md

@@ -1,18 +0,0 @@
----
-name: Technical Proposal
-about: Propose a new architecture or technology
-title: ''
-labels: 'proposal'
-assignees: ''
-
----
-
-**Summary**
-
-**Motivation**
-
-**Technical Design**
-
-**Alternatives to Consider**
-
-**Additional context**

.github/dependabot.yml

@@ -1,77 +0,0 @@
-version: 2
-updates:
-  - package-ecosystem: "pip"
-    directory: "/"
-    schedule:
-      interval: "daily"
-    open-pull-requests-limit: 1
-    groups:
-      # put packages in their own group if they have a history of breaking the build or needing to be reverted
-      pre-commit:
-        patterns:
-          - "pre-commit"
-      llama:
-        patterns:
-          - "llama*"
-      chromadb:
-        patterns:
-          - "chromadb"
-      browsergym:
-        patterns:
-          - "browsergym*"
-      security-all:
-        applies-to: "security-updates"
-        patterns:
-          - "*"
-      version-all:
-        applies-to: "version-updates"
-        patterns:
-          - "*"
-
-  - package-ecosystem: "npm"
-    directory: "/frontend"
-    schedule:
-      interval: "daily"
-    open-pull-requests-limit: 1
-    groups:
-      docusaurus:
-        patterns:
-          - "*docusaurus*"
-      eslint:
-        patterns:
-          - "*eslint*"
-      security-all:
-        applies-to: "security-updates"
-        patterns:
-          - "*"
-      version-all:
-        applies-to: "version-updates"
-        patterns:
-          - "*"
-
-  - package-ecosystem: "npm"
-    directory: "/docs"
-    schedule:
-      interval: "weekly"
-      day: "wednesday"
-    open-pull-requests-limit: 1
-    groups:
-      docusaurus:
-        patterns:
-          - "*docusaurus*"
-      eslint:
-        patterns:
-          - "*eslint*"
-      security-all:
-        applies-to: "security-updates"
-        patterns:
-          - "*"
-      version-all:
-        applies-to: "version-updates"
-        patterns:
-          - "*"
-
-  - package-ecosystem: "github-actions"
-    directory: "/"
-    schedule:
-      interval: "weekly"

.github/pull_request_template.md

@@ -1,12 +0,0 @@
-- [ ] This change is worth documenting at https://docs.all-hands.dev/
-- [ ] Include this change in the Release Notes. If checked, you **must** provide an **end-user friendly** description for your change below
-
-**End-user friendly description of the problem this fixes or functionality that this introduces.**
-
-
----
-**Give a summary of what the PR does, explaining any non-trivial design decisions.**
-
-
----
-**Link of any specific issues this addresses.**

.github/scripts/check_version_consistency.py

@@ -1,66 +0,0 @@
-#!/usr/bin/env python3
-import os
-import re
-import sys
-from typing import Set, Tuple
-
-
-def find_version_references(directory: str) -> Tuple[Set[str], Set[str]]:
-    openhands_versions = set()
-    runtime_versions = set()
-
-    version_pattern_openhands = re.compile(r'openhands:(\d{1})\.(\d{2})')
-    version_pattern_runtime = re.compile(r'runtime:(\d{1})\.(\d{2})')
-
-    for root, _, files in os.walk(directory):
-        # Skip .git directory
-        if '.git' in root:
-            continue
-
-        for file in files:
-            if file.endswith(
-                ('.md', '.yml', '.yaml', '.txt', '.html', '.py', '.js', '.ts')
-            ):
-                file_path = os.path.join(root, file)
-                try:
-                    with open(file_path, 'r', encoding='utf-8') as f:
-                        content = f.read()
-
-                        # Find all openhands version references
-                        matches = version_pattern_openhands.findall(content)
-                        openhands_versions.update(matches)
-
-                        # Find all runtime version references
-                        matches = version_pattern_runtime.findall(content)
-                        runtime_versions.update(matches)
-                except Exception as e:
-                    print(f'Error reading {file_path}: {e}', file=sys.stderr)
-
-    return openhands_versions, runtime_versions
-
-
-def main():
-    repo_root = os.path.abspath(os.path.join(os.path.dirname(__file__), '..', '..'))
-    openhands_versions, runtime_versions = find_version_references(repo_root)
-
-    exit_code = 0
-
-    if len(openhands_versions) > 1:
-        print('Error: Multiple openhands versions found:', file=sys.stderr)
-        print('Found versions:', sorted(openhands_versions), file=sys.stderr)
-        exit_code = 1
-    elif len(openhands_versions) == 0:
-        print('Warning: No openhands version references found', file=sys.stderr)
-
-    if len(runtime_versions) > 1:
-        print('Error: Multiple runtime versions found:', file=sys.stderr)
-        print('Found versions:', sorted(runtime_versions), file=sys.stderr)
-        exit_code = 1
-    elif len(runtime_versions) == 0:
-        print('Warning: No runtime version references found', file=sys.stderr)
-
-    sys.exit(exit_code)
-
-
-if __name__ == '__main__':
-    main()

.github/workflows/clean-up.yml

@@ -1,69 +0,0 @@
-# Workflow that cleans up outdated and old workflows to prevent out of disk issues
-name: Delete old workflow runs
-
-# This workflow is currently only triggered manually
-on:
-  workflow_dispatch:
-    inputs:
-      days:
-        description: 'Days-worth of runs to keep for each workflow'
-        required: true
-        default: '30'
-      minimum_runs:
-        description: 'Minimum runs to keep for each workflow'
-        required: true
-        default: '10'
-      delete_workflow_pattern:
-        description: 'Name or filename of the workflow (if not set, all workflows are targeted)'
-        required: false
-      delete_workflow_by_state_pattern:
-        description: 'Filter workflows by state: active, deleted, disabled_fork, disabled_inactivity, disabled_manually'
-        required: true
-        default: "ALL"
-        type: choice
-        options:
-          - "ALL"
-          - active
-          - deleted
-          - disabled_inactivity
-          - disabled_manually
-      delete_run_by_conclusion_pattern:
-        description: 'Remove runs based on conclusion: action_required, cancelled, failure, skipped, success'
-        required: true
-        default: 'ALL'
-        type: choice
-        options:
-          - 'ALL'
-          - 'Unsuccessful: action_required,cancelled,failure,skipped'
-          - action_required
-          - cancelled
-          - failure
-          - skipped
-          - success
-      dry_run:
-        description: 'Logs simulated changes, no deletions are performed'
-        required: false
-
-jobs:
-  del_runs:
-    runs-on: ubuntu-latest
-    permissions:
-      actions: write
-      contents: read
-    steps:
-      - name: Delete workflow runs
-        uses: Mattraks/delete-workflow-runs@v2
-        with:
-          token: ${{ github.token }}
-          repository: ${{ github.repository }}
-          retain_days: ${{ github.event.inputs.days }}
-          keep_minimum_runs: ${{ github.event.inputs.minimum_runs }}
-          delete_workflow_pattern: ${{ github.event.inputs.delete_workflow_pattern }}
-          delete_workflow_by_state_pattern: ${{ github.event.inputs.delete_workflow_by_state_pattern }}
-          delete_run_by_conclusion_pattern: >-
-            ${{
-              startsWith(github.event.inputs.delete_run_by_conclusion_pattern, 'Unsuccessful:')
-              && 'action_required,cancelled,failure,skipped'
-              || github.event.inputs.delete_run_by_conclusion_pattern
-            }}
-          dry_run: ${{ github.event.inputs.dry_run }}

.github/workflows/deploy-docs.yml

@@ -1,74 +0,0 @@
-# Workflow that builds and deploys the documentation website
-name: Deploy Docs to GitHub Pages
-
-# * Always run on "main"
-# * Run on PRs that target the "main" branch and have changes in the "docs" folder or this workflow
-on:
-  push:
-    branches:
-      - main
-  pull_request:
-    paths:
-      - 'docs/**'
-      - '.github/workflows/deploy-docs.yml'
-    branches:
-      - main
-
-# If triggered by a PR, it will be in the same group. However, each commit on main will be in its own unique group
-concurrency:
-  group: ${{ github.workflow }}-${{ (github.head_ref && github.ref) || github.run_id }}
-  cancel-in-progress: true
-
-jobs:
-  # Build the documentation website
-  build:
-    if: github.repository == 'All-Hands-AI/OpenHands'
-    name: Build Docusaurus
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-        with:
-          fetch-depth: 0
-      - uses: actions/setup-node@v4
-        with:
-          node-version: 18
-          cache: npm
-          cache-dependency-path: docs/package-lock.json
-      - name: Set up Python
-        uses: actions/setup-python@v5
-        with:
-          python-version: '3.12'
-      - name: Generate Python Docs
-        run: rm -rf docs/modules/python && pip install pydoc-markdown && pydoc-markdown
-      - name: Install dependencies
-        run: cd docs && npm ci
-      - name: Build website
-        run: cd docs && npm run build
-      - name: Upload Build Artifact
-        if: github.ref == 'refs/heads/main'
-        uses: actions/upload-pages-artifact@v3
-        with:
-          path: docs/build
-
-  # Deploy the documentation website
-  deploy:
-    if: github.ref == 'refs/heads/main' && github.repository == 'All-Hands-AI/OpenHands'
-    name: Deploy to GitHub Pages
-    runs-on: ubuntu-latest
-    # This job only runs on "main" so only run one of these jobs at a time
-    # otherwise it will fail if one is already running
-    concurrency:
-      group: ${{ github.workflow }}-${{ github.ref }}
-    needs: build
-    # Grant GITHUB_TOKEN the permissions required to make a Pages deployment
-    permissions:
-      pages: write # to deploy to Pages
-      id-token: write # to verify the deployment originates from an appropriate source
-    # Deploy to the github-pages environment
-    environment:
-      name: github-pages
-      url: ${{ steps.deployment.outputs.page_url }}
-    steps:
-      - name: Deploy to GitHub Pages
-        id: deployment
-        uses: actions/deploy-pages@v4

.github/workflows/dummy-agent-test.yml

@@ -1,53 +0,0 @@
-# Workflow that uses the DummyAgent to run a simple task
-name: Run E2E test with dummy agent
-
-# Always run on "main"
-# Always run on PRs
-on:
-  push:
-    branches:
-    - main
-  pull_request:
-
-# If triggered by a PR, it will be in the same group. However, each commit on main will be in its own unique group
-concurrency:
-  group: ${{ github.workflow }}-${{ (github.head_ref && github.ref) || github.run_id }}
-  cancel-in-progress: true
-
-jobs:
-  test:
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-      - name: Set up Docker Buildx
-        id: buildx
-        uses: docker/setup-buildx-action@v3
-      - name: Install tmux
-        run: sudo apt-get update && sudo apt-get install -y tmux
-      - name: Setup Node.js
-        uses: actions/setup-node@v4
-        with:
-          node-version: '22.x'
-      - name: Install poetry via pipx
-        run: pipx install poetry
-      - name: Set up Python
-        uses: actions/setup-python@v5
-        with:
-          python-version: '3.12'
-          cache: 'poetry'
-      - name: Install Python dependencies using Poetry
-        run: poetry install --without evaluation,llama-index
-      - name: Build Environment
-        run: make build
-      - name: Run tests
-        run: |
-          set -e
-          SANDBOX_FORCE_REBUILD_RUNTIME=True poetry run python3 openhands/core/main.py -t "do a flip" -d ./workspace/ -c DummyAgent
-      - name: Check exit code
-        run: |
-          if [ $? -ne 0 ]; then
-            echo "Test failed"
-            exit 1
-          else
-            echo "Test passed"
-          fi

.github/workflows/eval-runner.yml

@@ -1,139 +0,0 @@
-name: Run SWE-Bench Evaluation
-
-on:
-  pull_request:
-    types: [labeled]
-  workflow_dispatch:
-    inputs:
-      reason:
-        description: "Reason for manual trigger"
-        required: true
-        default: ""
-
-env:
-  N_PROCESSES: 32 # Global configuration for number of parallel processes for evaluation
-
-jobs:
-  run-evaluation:
-    if: github.event.label.name == 'eval-this' || github.event_name != 'pull_request'
-    runs-on: ubuntu-latest
-    permissions:
-      contents: "read"
-      id-token: "write"
-      pull-requests: "write"
-      issues: "write"
-    strategy:
-      matrix:
-        python-version: ["3.12"]
-    steps:
-      - name: Checkout repository
-        uses: actions/checkout@v4
-
-      - name: Install tmux
-        run: sudo apt-get update && sudo apt-get install -y tmux
-      - name: Install poetry via pipx
-        run: pipx install poetry
-
-      - name: Set up Python
-        uses: actions/setup-python@v5
-        with:
-          python-version: ${{ matrix.python-version }}
-          cache: "poetry"
-
-      - name: Comment on PR if 'eval-this' label is present
-        if: github.event_name == 'pull_request' && github.event.label.name == 'eval-this'
-        uses: KeisukeYamashita/create-comment@v1
-        with:
-          unique: false
-          comment: |
-            Hi! I started running the evaluation on your PR. You will receive a comment with the results shortly.
-
-      - name: Install Python dependencies using Poetry
-        run: poetry install
-
-      - name: Configure config.toml for evaluation
-        env:
-          DEEPSEEK_API_KEY: ${{ secrets.DEEPSEEK_LLM_API_KEY }}
-        run: |
-          echo "[llm.eval]" > config.toml
-          echo "model = \"deepseek/deepseek-chat\"" >> config.toml
-          echo "api_key = \"$DEEPSEEK_API_KEY\"" >> config.toml
-          echo "temperature = 0.0" >> config.toml
-
-      - name: Run SWE-Bench evaluation
-        env:
-          ALLHANDS_API_KEY: ${{ secrets.ALLHANDS_EVAL_RUNTIME_API_KEY }}
-          RUNTIME: remote
-          SANDBOX_REMOTE_RUNTIME_API_URL: https://runtime.eval.all-hands.dev
-          EVAL_DOCKER_IMAGE_PREFIX: us-central1-docker.pkg.dev/evaluation-092424/swe-bench-images
-
-        run: |
-          poetry run ./evaluation/benchmarks/swe_bench/scripts/run_infer.sh llm.eval HEAD CodeActAgent 300 30 $N_PROCESSES "princeton-nlp/SWE-bench_Lite" test
-          OUTPUT_FOLDER=$(find evaluation/evaluation_outputs/outputs/princeton-nlp__SWE-bench_Lite-test/CodeActAgent -name "deepseek-chat_maxiter_50_N_*-no-hint-run_1" -type d | head -n 1)
-          echo "OUTPUT_FOLDER for SWE-bench evaluation: $OUTPUT_FOLDER"
-          poetry run ./evaluation/benchmarks/swe_bench/scripts/eval_infer_remote.sh $OUTPUT_FOLDER/output.jsonl $N_PROCESSES "princeton-nlp/SWE-bench_Lite" test
-
-          poetry run ./evaluation/benchmarks/swe_bench/scripts/eval/summarize_outputs.py $OUTPUT_FOLDER/output.jsonl > summarize_outputs.log 2>&1
-          echo "SWEBENCH_REPORT<<EOF" >> $GITHUB_ENV
-          cat summarize_outputs.log >> $GITHUB_ENV
-          echo "EOF" >> $GITHUB_ENV
-
-      - name: Create tar.gz of evaluation outputs
-        run: |
-          TIMESTAMP=$(date +'%y-%m-%d-%H-%M')
-          tar -czvf evaluation_outputs_${TIMESTAMP}.tar.gz evaluation/evaluation_outputs/outputs
-
-      - name: Upload evaluation results as artifact
-        uses: actions/upload-artifact@v4
-        id: upload_results_artifact
-        with:
-          name: evaluation-outputs
-          path: evaluation_outputs_*.tar.gz
-
-      - name: Get artifact URL
-        run: echo "ARTIFACT_URL=${{ steps.upload_results_artifact.outputs.artifact-url }}" >> $GITHUB_ENV
-
-      - name: Authenticate to Google Cloud
-        uses: 'google-github-actions/auth@v2'
-        with:
-          credentials_json: ${{ secrets.GCP_RESEARCH_OBJECT_CREATOR_SA_KEY }}
-
-      - name: Set timestamp and trigger reason
-        run: |
-          echo "TIMESTAMP=$(date +'%Y-%m-%d-%H-%M')" >> $GITHUB_ENV
-          if [[ "${{ github.event_name }}" == "pull_request" ]]; then
-            echo "TRIGGER_REASON=pr-${{ github.event.pull_request.number }}" >> $GITHUB_ENV
-          elif [[ "${{ github.event_name }}" == "schedule" ]]; then
-            echo "TRIGGER_REASON=schedule" >> $GITHUB_ENV
-          else
-            echo "TRIGGER_REASON=manual-${{ github.event.inputs.reason }}" >> $GITHUB_ENV
-          fi
-
-      - name: Upload evaluation results to Google Cloud Storage
-        uses: 'google-github-actions/upload-cloud-storage@v2'
-        with:
-          path: 'evaluation/evaluation_outputs/outputs'
-          destination: 'openhands-oss-eval-results/${{ env.TIMESTAMP }}-${{ env.TRIGGER_REASON }}'
-
-      - name: Comment with evaluation results and artifact link
-        id: create_comment
-        uses: KeisukeYamashita/create-comment@v1
-        with:
-          number: ${{ github.event_name == 'pull_request' && github.event.pull_request.number || 4504 }}
-          unique: false
-          comment: |
-              Trigger by: ${{ github.event_name == 'pull_request' && format('Pull Request (eval-this label on PR #{0})', github.event.pull_request.number) || github.event_name == 'schedule' && 'Daily Schedule' || format('Manual Trigger: {0}', github.event.inputs.reason) }}
-              Commit: ${{ github.sha }}
-              **SWE-Bench Evaluation Report**
-              ${{ env.SWEBENCH_REPORT }}
-              ---
-              You can download the full evaluation outputs [here](${{ env.ARTIFACT_URL }}).
-
-      - name: Post to a Slack channel
-        id: slack
-        uses: slackapi/slack-github-action@v2.0.0
-        with:
-          channel-id: 'C07SVQSCR6F'
-          slack-message: "*Evaluation Trigger:* ${{ github.event_name == 'pull_request' && format('Pull Request (eval-this label on PR #{0})', github.event.pull_request.number) || github.event_name == 'schedule' && 'Daily Schedule' || format('Manual Trigger: {0}', github.event.inputs.reason) }}\n\nLink to summary: [here](https://github.com/${{ github.repository }}/issues/${{ github.event_name == 'pull_request' && github.event.pull_request.number || 4504 }}#issuecomment-${{ steps.create_comment.outputs.comment-id }})"
-        env:
-          SLACK_BOT_TOKEN: ${{ secrets.EVAL_NOTIF_SLACK_BOT_TOKEN }}

.github/workflows/fe-unit-tests.yml

@@ -1,48 +0,0 @@
-# Workflow that runs frontend unit tests
-name: Run Frontend Unit Tests
-
-# * Always run on "main"
-# * Run on PRs that have changes in the "frontend" folder or this workflow
-on:
-  push:
-    branches:
-      - main
-  pull_request:
-    paths:
-      - 'frontend/**'
-      -  '.github/workflows/fe-unit-tests.yml'
-
-# If triggered by a PR, it will be in the same group. However, each commit on main will be in its own unique group
-concurrency:
-  group: ${{ github.workflow }}-${{ (github.head_ref && github.ref) || github.run_id }}
-  cancel-in-progress: true
-
-jobs:
-  # Run frontend unit tests
-  fe-test:
-    name: FE Unit Tests
-    runs-on: ubuntu-latest
-    strategy:
-      matrix:
-        node-version: [20, 22]
-      fail-fast: true
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-      - name: Set up Node.js
-        uses: actions/setup-node@v4
-        with:
-          node-version: ${{ matrix.node-version }}
-      - name: Install dependencies
-        working-directory: ./frontend
-        run: npm ci
-      - name: Run TypeScript compilation
-        working-directory: ./frontend
-        run: npm run make-i18n && tsc
-      - name: Run tests and collect coverage
-        working-directory: ./frontend
-        run: npm run test:coverage
-      - name: Upload coverage to Codecov
-        uses: codecov/codecov-action@v5
-        env:
-          CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}

.github/workflows/ghcr-build.yml

@@ -1,396 +0,0 @@
-# Workflow that builds, tests and then pushes the OpenHands and runtime docker images to the ghcr.io repository
-name: Docker
-
-# Always run on "main"
-# Always run on tags
-# Always run on PRs
-# Can also be triggered manually
-on:
-  push:
-    branches:
-      - main
-    tags:
-      - '*'
-  pull_request:
-  workflow_dispatch:
-    inputs:
-      reason:
-        description: 'Reason for manual trigger'
-        required: true
-        default: ''
-
-# If triggered by a PR, it will be in the same group. However, each commit on main will be in its own unique group
-concurrency:
-  group: ${{ github.workflow }}-${{ (github.head_ref && github.ref) || github.run_id }}
-  cancel-in-progress: true
-
-env:
-  BASE_IMAGE_FOR_HASH_EQUIVALENCE_TEST: nikolaik/python-nodejs:python3.12-nodejs22
-  RELEVANT_SHA: ${{ github.event.pull_request.head.sha || github.sha }}
-
-jobs:
-  # Builds the OpenHands Docker images
-  ghcr_build_app:
-    name: Build App Image
-    runs-on: ubuntu-latest
-    permissions:
-      contents: read
-      packages: write
-    outputs:
-      hash_from_app_image: ${{ steps.get_hash_in_app_image.outputs.hash_from_app_image }}
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-      - name: Set up QEMU
-        uses: docker/setup-qemu-action@v3.4.0
-        with:
-          image: tonistiigi/binfmt:latest
-      - name: Login to GHCR
-        uses: docker/login-action@v3
-        with:
-          registry: ghcr.io
-          username: ${{ github.repository_owner }}
-          password: ${{ secrets.GITHUB_TOKEN }}
-      - name: Set up Docker Buildx
-        id: buildx
-        uses: docker/setup-buildx-action@v3
-      - name: Lowercase Repository Owner
-        run: |
-          echo REPO_OWNER=$(echo ${{ github.repository_owner }} | tr '[:upper:]' '[:lower:]') >> $GITHUB_ENV
-      - name: Build and push app image
-        if: "!github.event.pull_request.head.repo.fork"
-        run: |
-          ./containers/build.sh -i openhands -o ${{ env.REPO_OWNER }} --push
-      - name: Build app image
-        if: "github.event.pull_request.head.repo.fork"
-        run: |
-          ./containers/build.sh -i openhands -o ${{ env.REPO_OWNER }} --load
-      - name: Get hash in App Image
-        id: get_hash_in_app_image
-        run: |
-          # Run the build script in the app image
-          docker run -e SANDBOX_USER_ID=0 -v /var/run/docker.sock:/var/run/docker.sock ghcr.io/${{ env.REPO_OWNER }}/openhands:${{ env.RELEVANT_SHA }} /bin/bash -c "mkdir -p containers/runtime; python3 openhands/runtime/utils/runtime_build.py --base_image ${{ env.BASE_IMAGE_FOR_HASH_EQUIVALENCE_TEST }} --build_folder containers/runtime --force_rebuild" 2>&1 | tee docker-outputs.txt
-          # Get the hash from the build script
-          hash_from_app_image=$(cat docker-outputs.txt | grep "Hash for docker build directory" | awk -F "): " '{print $2}' | uniq | head -n1)
-          echo "hash_from_app_image=$hash_from_app_image" >> $GITHUB_OUTPUT
-          echo "Hash from app image: $hash_from_app_image"
-
-  # Builds the runtime Docker images
-  ghcr_build_runtime:
-    name: Build Image
-    runs-on: ubuntu-latest
-    permissions:
-      contents: read
-      packages: write
-    strategy:
-      matrix:
-        base_image:
-          - image: 'nikolaik/python-nodejs:python3.12-nodejs22'
-            tag: nikolaik
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-      - name: Set up QEMU
-        uses: docker/setup-qemu-action@v3.4.0
-        with:
-          image: tonistiigi/binfmt:latest
-      - name: Login to GHCR
-        uses: docker/login-action@v3
-        with:
-          registry: ghcr.io
-          username: ${{ github.repository_owner }}
-          password: ${{ secrets.GITHUB_TOKEN }}
-      - name: Set up Docker Buildx
-        id: buildx
-        uses: docker/setup-buildx-action@v3
-      - name: Set up Python
-        uses: actions/setup-python@v5
-        with:
-          python-version: '3.12'
-      - name: Cache Poetry dependencies
-        uses: actions/cache@v4
-        with:
-          path: |
-            ~/.cache/pypoetry
-            ~/.virtualenvs
-          key: ${{ runner.os }}-poetry-${{ hashFiles('**/poetry.lock') }}
-          restore-keys: |
-            ${{ runner.os }}-poetry-
-      - name: Install poetry via pipx
-        run: pipx install poetry
-      - name: Install Python dependencies using Poetry
-        run: make install-python-dependencies
-      - name: Create source distribution and Dockerfile
-        run: poetry run python3 openhands/runtime/utils/runtime_build.py --base_image ${{ matrix.base_image.image }} --build_folder containers/runtime --force_rebuild
-      - name: Lowercase Repository Owner
-        run: |
-          echo REPO_OWNER=$(echo ${{ github.repository_owner }} | tr '[:upper:]' '[:lower:]') >> $GITHUB_ENV
-      - name: Build and push runtime image ${{ matrix.base_image.image }}
-        if: github.event.pull_request.head.repo.fork != true
-        run: |
-          ./containers/build.sh -i runtime -o ${{ env.REPO_OWNER }} --push -t ${{ matrix.base_image.tag }}
-      # Forked repos can't push to GHCR, so we need to upload the image as an artifact
-      - name: Build runtime image ${{ matrix.base_image.image }} for fork
-        if: github.event.pull_request.head.repo.fork
-        uses: docker/build-push-action@v6
-        with:
-          tags: ghcr.io/${{ env.REPO_OWNER }}/runtime:${{ env.RELEVANT_SHA }}-${{ matrix.base_image.tag }}
-          outputs: type=docker,dest=/tmp/runtime-${{ matrix.base_image.tag }}.tar
-          context: containers/runtime
-      - name: Upload runtime image for fork
-        if: github.event.pull_request.head.repo.fork
-        uses: actions/upload-artifact@v4
-        with:
-          name: runtime-${{ matrix.base_image.tag }}
-          path: /tmp/runtime-${{ matrix.base_image.tag }}.tar
-
-  verify_hash_equivalence_in_runtime_and_app:
-    name: Verify Hash Equivalence in Runtime and Docker images
-    runs-on: ubuntu-latest
-    needs: [ghcr_build_runtime, ghcr_build_app]
-    strategy:
-      fail-fast: false
-      matrix:
-        base_image: ['nikolaik']
-    steps:
-      - uses: actions/checkout@v4
-      - name: Cache Poetry dependencies
-        uses: actions/cache@v4
-        with:
-          path: |
-            ~/.cache/pypoetry
-            ~/.virtualenvs
-          key: ${{ runner.os }}-poetry-${{ hashFiles('**/poetry.lock') }}
-          restore-keys: |
-            ${{ runner.os }}-poetry-
-      - name: Set up Python
-        uses: actions/setup-python@v5
-        with:
-          python-version: '3.12'
-      - name: Install poetry via pipx
-        run: pipx install poetry
-      - name: Install Python dependencies using Poetry
-        run: make install-python-dependencies
-      - name: Get hash in App Image
-        run: |
-          echo "Hash from app image: ${{ needs.ghcr_build_app.outputs.hash_from_app_image }}"
-          echo "hash_from_app_image=${{ needs.ghcr_build_app.outputs.hash_from_app_image }}" >> $GITHUB_ENV
-
-      - name: Get hash using code (development mode)
-        run: |
-          mkdir -p containers/runtime
-          poetry run python3 openhands/runtime/utils/runtime_build.py --base_image ${{ env.BASE_IMAGE_FOR_HASH_EQUIVALENCE_TEST }} --build_folder containers/runtime --force_rebuild > output.txt 2>&1
-          hash_from_code=$(cat output.txt | grep "Hash for docker build directory" | awk -F "): " '{print $2}' | uniq | head -n1)
-          echo "hash_from_code=$hash_from_code" >> $GITHUB_ENV
-
-      - name: Compare hashes
-        run: |
-          echo "Hash from App Image: ${{ env.hash_from_app_image }}"
-          echo "Hash from Code: ${{ env.hash_from_code }}"
-          if [ "${{ env.hash_from_app_image }}" = "${{ env.hash_from_code }}" ]; then
-            echo "Hashes match!"
-          else
-            echo "Hashes do not match!"
-            exit 1
-          fi
-
-  # Run unit tests with the Docker runtime Docker images as root
-  test_runtime_root:
-    name: RT Unit Tests (Root)
-    needs: [ghcr_build_runtime]
-    runs-on: ubuntu-latest
-    strategy:
-      fail-fast: false
-      matrix:
-        base_image: ['nikolaik']
-    steps:
-      - uses: actions/checkout@v4
-      - name: Set up Docker Buildx
-        id: buildx
-        uses: docker/setup-buildx-action@v3
-      # Forked repos can't push to GHCR, so we need to download the image as an artifact
-      - name: Download runtime image for fork
-        if: github.event.pull_request.head.repo.fork
-        uses: actions/download-artifact@v4
-        with:
-          name: runtime-${{ matrix.base_image }}
-          path: /tmp
-      - name: Load runtime image for fork
-        if: github.event.pull_request.head.repo.fork
-        run: |
-          docker load --input /tmp/runtime-${{ matrix.base_image }}.tar
-      - name: Cache Poetry dependencies
-        uses: actions/cache@v4
-        with:
-          path: |
-            ~/.cache/pypoetry
-            ~/.virtualenvs
-          key: ${{ runner.os }}-poetry-${{ hashFiles('**/poetry.lock') }}
-          restore-keys: |
-            ${{ runner.os }}-poetry-
-      - name: Set up Python
-        uses: actions/setup-python@v5
-        with:
-          python-version: '3.12'
-      - name: Install poetry via pipx
-        run: pipx install poetry
-      - name: Install Python dependencies using Poetry
-        run: make install-python-dependencies
-      - name: Lowercase Repository Owner
-        run: |
-          echo REPO_OWNER=$(echo ${{ github.repository_owner }} | tr '[:upper:]' '[:lower:]') >> $GITHUB_ENV
-      - name: Run docker runtime tests
-        run: |
-          # We install pytest-xdist in order to run tests across CPUs
-          poetry run pip install pytest-xdist
-
-          # Install to be able to retry on failures for flaky tests
-          poetry run pip install pytest-rerunfailures
-
-          image_name=ghcr.io/${{ env.REPO_OWNER }}/runtime:${{ env.RELEVANT_SHA }}-${{ matrix.base_image }}
-
-          TEST_RUNTIME=docker \
-          SANDBOX_USER_ID=$(id -u) \
-          SANDBOX_RUNTIME_CONTAINER_IMAGE=$image_name \
-          TEST_IN_CI=true \
-          RUN_AS_OPENHANDS=false \
-          poetry run pytest -n 3 -raRs --reruns 2 --reruns-delay 5 --cov=openhands --cov-report=xml -s ./tests/runtime --ignore=tests/runtime/test_browsergym_envs.py
-      - name: Upload coverage to Codecov
-        uses: codecov/codecov-action@v5
-        env:
-          CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
-
-  # Run unit tests with the Docker runtime Docker images as openhands user
-  test_runtime_oh:
-    name: RT Unit Tests (openhands)
-    runs-on: ubuntu-latest
-    needs: [ghcr_build_runtime]
-    strategy:
-      matrix:
-        base_image: ['nikolaik']
-    steps:
-      - uses: actions/checkout@v4
-      - name: Set up Docker Buildx
-        id: buildx
-        uses: docker/setup-buildx-action@v3
-      # Forked repos can't push to GHCR, so we need to download the image as an artifact
-      - name: Download runtime image for fork
-        if: github.event.pull_request.head.repo.fork
-        uses: actions/download-artifact@v4
-        with:
-          name: runtime-${{ matrix.base_image }}
-          path: /tmp
-      - name: Load runtime image for fork
-        if: github.event.pull_request.head.repo.fork
-        run: |
-          docker load --input /tmp/runtime-${{ matrix.base_image }}.tar
-      - name: Cache Poetry dependencies
-        uses: actions/cache@v4
-        with:
-          path: |
-            ~/.cache/pypoetry
-            ~/.virtualenvs
-          key: ${{ runner.os }}-poetry-${{ hashFiles('**/poetry.lock') }}
-          restore-keys: |
-            ${{ runner.os }}-poetry-
-      - name: Set up Python
-        uses: actions/setup-python@v5
-        with:
-          python-version: '3.12'
-      - name: Install poetry via pipx
-        run: pipx install poetry
-      - name: Install Python dependencies using Poetry
-        run: make install-python-dependencies
-      - name: Lowercase Repository Owner
-        run: |
-          echo REPO_OWNER=$(echo ${{ github.repository_owner }} | tr '[:upper:]' '[:lower:]') >> $GITHUB_ENV
-      - name: Run runtime tests
-        run: |
-          # We install pytest-xdist in order to run tests across CPUs
-          poetry run pip install pytest-xdist
-
-          # Install to be able to retry on failures for flaky tests
-          poetry run pip install pytest-rerunfailures
-
-          image_name=ghcr.io/${{ env.REPO_OWNER }}/runtime:${{ env.RELEVANT_SHA }}-${{ matrix.base_image }}
-
-          TEST_RUNTIME=docker \
-          SANDBOX_USER_ID=$(id -u) \
-          SANDBOX_RUNTIME_CONTAINER_IMAGE=$image_name \
-          TEST_IN_CI=true \
-          RUN_AS_OPENHANDS=true \
-          poetry run pytest -n 3 -raRs --reruns 2 --reruns-delay 5 --cov=openhands --cov-report=xml -s ./tests/runtime --ignore=tests/runtime/test_browsergym_envs.py
-      - name: Upload coverage to Codecov
-        uses: codecov/codecov-action@v5
-        env:
-          CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
-
-  # The two following jobs (named identically) are to check whether all the runtime tests have passed as the
-  # "All Runtime Tests Passed" is a required job for PRs to merge
-  # Due to this bug: https://github.com/actions/runner/issues/2566, we want to create a job that runs when the
-  # prerequisites have been cancelled or failed so merging is disallowed, otherwise Github considers "skipped" as "success"
-  runtime_tests_check_success:
-    name: All Runtime Tests Passed
-    if: ${{ !cancelled() && !contains(needs.*.result, 'failure') && !contains(needs.*.result, 'cancelled') }}
-    runs-on: ubuntu-latest
-    needs: [test_runtime_root, test_runtime_oh, verify_hash_equivalence_in_runtime_and_app]
-    steps:
-      - name: All tests passed
-        run: echo "All runtime tests have passed successfully!"
-
-  runtime_tests_check_fail:
-    name: All Runtime Tests Passed
-    if: ${{ cancelled() || contains(needs.*.result, 'failure') || contains(needs.*.result, 'cancelled') }}
-    runs-on: ubuntu-latest
-    needs: [test_runtime_root, test_runtime_oh, verify_hash_equivalence_in_runtime_and_app]
-    steps:
-      - name: Some tests failed
-        run: |
-          echo "Some runtime tests failed or were cancelled"
-          exit 1
-  update_pr_description:
-    name: Update PR Description
-    if: github.event_name == 'pull_request' && !github.event.pull_request.head.repo.fork && github.actor != 'dependabot[bot]'
-    needs: [ghcr_build_runtime]
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-
-      - name: Get short SHA
-        id: short_sha
-        run: echo "SHORT_SHA=$(echo ${{ github.event.pull_request.head.sha }} | cut -c1-7)" >> $GITHUB_OUTPUT
-
-      - name: Update PR Description
-        env:
-          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-          PR_NUMBER: ${{ github.event.pull_request.number }}
-          REPO: ${{ github.repository }}
-          SHORT_SHA: ${{ steps.short_sha.outputs.SHORT_SHA }}
-        run: |
-          echo "updating PR description"
-          DOCKER_RUN_COMMAND="docker run -it --rm \
-            -p 3000:3000 \
-            -v /var/run/docker.sock:/var/run/docker.sock \
-            --add-host host.docker.internal:host-gateway \
-            -e SANDBOX_RUNTIME_CONTAINER_IMAGE=docker.all-hands.dev/all-hands-ai/runtime:$SHORT_SHA-nikolaik \
-            --name openhands-app-$SHORT_SHA \
-            docker.all-hands.dev/all-hands-ai/openhands:$SHORT_SHA"
-
-          PR_BODY=$(gh pr view $PR_NUMBER --json body --jq .body)
-
-          if echo "$PR_BODY" | grep -q "To run this PR locally, use the following command:"; then
-            UPDATED_PR_BODY=$(echo "${PR_BODY}" | sed -E "s|docker run -it --rm.*|$DOCKER_RUN_COMMAND|")
-          else
-            UPDATED_PR_BODY="${PR_BODY}
-
-          ---
-
-          To run this PR locally, use the following command:
-          \`\`\`
-          $DOCKER_RUN_COMMAND
-          \`\`\`"
-          fi
-
-          echo "updated body: $UPDATED_PR_BODY"
-          gh pr edit $PR_NUMBER --body "$UPDATED_PR_BODY"

.github/workflows/integration-runner.yml

@@ -1,267 +0,0 @@
-name: Run Integration Tests
-
-on:
-  pull_request:
-    types: [labeled]
-  workflow_dispatch:
-    inputs:
-      reason:
-        description: 'Reason for manual trigger'
-        required: true
-        default: ''
-  schedule:
-    - cron: '30 22 * * *'  # Runs at 10:30pm UTC every day
-
-env:
-  N_PROCESSES: 10 # Global configuration for number of parallel processes for evaluation
-
-jobs:
-  run-integration-tests:
-    if: github.event.label.name == 'integration-test' || github.event_name == 'workflow_dispatch' || github.event_name == 'schedule'
-    runs-on: ubuntu-latest
-    permissions:
-      contents: "read"
-      id-token: "write"
-      pull-requests: "write"
-      issues: "write"
-    strategy:
-      matrix:
-        python-version: ["3.12"]
-    steps:
-      - name: Checkout repository
-        uses: actions/checkout@v4
-
-      - name: Install poetry via pipx
-        run: pipx install poetry
-
-      - name: Set up Python
-        uses: actions/setup-python@v5
-        with:
-          python-version: ${{ matrix.python-version }}
-          cache: "poetry"
-
-      - name: Setup Node.js
-        uses: actions/setup-node@v4
-        with:
-          node-version: '22.x'
-
-      - name: Comment on PR if 'integration-test' label is present
-        if: github.event_name == 'pull_request' && github.event.label.name == 'integration-test'
-        uses: KeisukeYamashita/create-comment@v1
-        with:
-          unique: false
-          comment: |
-            Hi! I started running the integration tests on your PR. You will receive a comment with the results shortly.
-
-      - name: Install Python dependencies using Poetry
-        run: poetry install --without evaluation,llama-index
-
-      - name: Configure config.toml for testing with Haiku
-        env:
-          LLM_MODEL: "litellm_proxy/claude-3-5-haiku-20241022"
-          LLM_API_KEY: ${{ secrets.LLM_API_KEY }}
-          LLM_BASE_URL: ${{ secrets.LLM_BASE_URL }}
-          MAX_ITERATIONS: 10
-        run: |
-          echo "[llm.eval]" > config.toml
-          echo "model = \"$LLM_MODEL\"" >> config.toml
-          echo "api_key = \"$LLM_API_KEY\"" >> config.toml
-          echo "base_url = \"$LLM_BASE_URL\"" >> config.toml
-          echo "temperature = 0.0" >> config.toml
-
-      - name: Build environment
-        run: make build
-
-      - name: Run integration test evaluation for Haiku
-        env:
-          SANDBOX_FORCE_REBUILD_RUNTIME: True
-        run: |
-          poetry run ./evaluation/integration_tests/scripts/run_infer.sh llm.eval HEAD CodeActAgent '' 10 $N_PROCESSES '' 'haiku_run'
-
-          # get integration tests report
-          REPORT_FILE_HAIKU=$(find evaluation/evaluation_outputs/outputs/integration_tests/CodeActAgent/*haiku*_maxiter_10_N* -name "report.md" -type f | head -n 1)
-          echo "REPORT_FILE: $REPORT_FILE_HAIKU"
-          echo "INTEGRATION_TEST_REPORT_HAIKU<<EOF" >> $GITHUB_ENV
-          cat $REPORT_FILE_HAIKU >> $GITHUB_ENV
-          echo >> $GITHUB_ENV
-          echo "EOF" >> $GITHUB_ENV
-
-      - name: Wait a little bit
-        run: sleep 10
-
-      - name: Configure config.toml for testing with DeepSeek
-        env:
-          LLM_MODEL: "litellm_proxy/deepseek-chat"
-          LLM_API_KEY: ${{ secrets.LLM_API_KEY }}
-          LLM_BASE_URL: ${{ secrets.LLM_BASE_URL }}
-          MAX_ITERATIONS: 10
-        run: |
-          echo "[llm.eval]" > config.toml
-          echo "model = \"$LLM_MODEL\"" >> config.toml
-          echo "api_key = \"$LLM_API_KEY\"" >> config.toml
-          echo "base_url = \"$LLM_BASE_URL\"" >> config.toml
-          echo "temperature = 0.0" >> config.toml
-
-      - name: Run integration test evaluation for DeepSeek
-        env:
-          SANDBOX_FORCE_REBUILD_RUNTIME: True
-        run: |
-          poetry run ./evaluation/integration_tests/scripts/run_infer.sh llm.eval HEAD CodeActAgent '' 10 $N_PROCESSES '' 'deepseek_run'
-
-          # get integration tests report
-          REPORT_FILE_DEEPSEEK=$(find evaluation/evaluation_outputs/outputs/integration_tests/CodeActAgent/deepseek*_maxiter_10_N* -name "report.md" -type f | head -n 1)
-          echo "REPORT_FILE: $REPORT_FILE_DEEPSEEK"
-          echo "INTEGRATION_TEST_REPORT_DEEPSEEK<<EOF" >> $GITHUB_ENV
-          cat $REPORT_FILE_DEEPSEEK >> $GITHUB_ENV
-          echo >> $GITHUB_ENV
-          echo "EOF" >> $GITHUB_ENV
-
-      # -------------------------------------------------------------
-      # Run DelegatorAgent tests for Haiku, limited to t01 and t02
-      - name: Wait a little bit (again)
-        run: sleep 5
-
-      - name: Configure config.toml for testing DelegatorAgent (Haiku)
-        env:
-          LLM_MODEL: "litellm_proxy/claude-3-5-haiku-20241022"
-          LLM_API_KEY: ${{ secrets.LLM_API_KEY }}
-          LLM_BASE_URL: ${{ secrets.LLM_BASE_URL }}
-          MAX_ITERATIONS: 30
-        run: |
-          echo "[llm.eval]" > config.toml
-          echo "model = \"$LLM_MODEL\"" >> config.toml
-          echo "api_key = \"$LLM_API_KEY\"" >> config.toml
-          echo "base_url = \"$LLM_BASE_URL\"" >> config.toml
-          echo "temperature = 0.0" >> config.toml
-
-      - name: Run integration test evaluation for DelegatorAgent (Haiku)
-        env:
-          SANDBOX_FORCE_REBUILD_RUNTIME: True
-        run: |
-          poetry run ./evaluation/integration_tests/scripts/run_infer.sh llm.eval HEAD DelegatorAgent '' 30 $N_PROCESSES "t01_fix_simple_typo,t02_add_bash_hello" 'delegator_haiku_run'
-
-          # Find and export the delegator test results
-          REPORT_FILE_DELEGATOR_HAIKU=$(find evaluation/evaluation_outputs/outputs/integration_tests/DelegatorAgent/*haiku*_maxiter_30_N* -name "report.md" -type f | head -n 1)
-          echo "REPORT_FILE_DELEGATOR_HAIKU: $REPORT_FILE_DELEGATOR_HAIKU"
-          echo "INTEGRATION_TEST_REPORT_DELEGATOR_HAIKU<<EOF" >> $GITHUB_ENV
-          cat $REPORT_FILE_DELEGATOR_HAIKU >> $GITHUB_ENV
-          echo >> $GITHUB_ENV
-          echo "EOF" >> $GITHUB_ENV
-
-      # -------------------------------------------------------------
-      # Run DelegatorAgent tests for DeepSeek, limited to t01 and t02
-      - name: Wait a little bit (again)
-        run: sleep 5
-
-      - name: Configure config.toml for testing DelegatorAgent (DeepSeek)
-        env:
-          LLM_MODEL: "litellm_proxy/deepseek-chat"
-          LLM_API_KEY: ${{ secrets.LLM_API_KEY }}
-          LLM_BASE_URL: ${{ secrets.LLM_BASE_URL }}
-          MAX_ITERATIONS: 30
-        run: |
-          echo "[llm.eval]" > config.toml
-          echo "model = \"$LLM_MODEL\"" >> config.toml
-          echo "api_key = \"$LLM_API_KEY\"" >> config.toml
-          echo "base_url = \"$LLM_BASE_URL\"" >> config.toml
-          echo "temperature = 0.0" >> config.toml
-      - name: Run integration test evaluation for DelegatorAgent (DeepSeek)
-        env:
-          SANDBOX_FORCE_REBUILD_RUNTIME: True
-        run: |
-          poetry run ./evaluation/integration_tests/scripts/run_infer.sh llm.eval HEAD DelegatorAgent '' 30 $N_PROCESSES "t01_fix_simple_typo,t02_add_bash_hello" 'delegator_deepseek_run'
-
-          # Find and export the delegator test results
-          REPORT_FILE_DELEGATOR_DEEPSEEK=$(find evaluation/evaluation_outputs/outputs/integration_tests/DelegatorAgent/deepseek*_maxiter_30_N* -name "report.md" -type f | head -n 1)
-          echo "REPORT_FILE_DELEGATOR_DEEPSEEK: $REPORT_FILE_DELEGATOR_DEEPSEEK"
-          echo "INTEGRATION_TEST_REPORT_DELEGATOR_DEEPSEEK<<EOF" >> $GITHUB_ENV
-          cat $REPORT_FILE_DELEGATOR_DEEPSEEK >> $GITHUB_ENV
-          echo >> $GITHUB_ENV
-          echo "EOF" >> $GITHUB_ENV
-      # -------------------------------------------------------------
-      # Run VisualBrowsingAgent tests for DeepSeek, limited to t05 and t06
-      - name: Wait a little bit (again)
-        run: sleep 5
-
-      - name: Configure config.toml for testing VisualBrowsingAgent (DeepSeek)
-        env:
-          LLM_MODEL: "litellm_proxy/deepseek-chat"
-          LLM_API_KEY: ${{ secrets.LLM_API_KEY }}
-          LLM_BASE_URL: ${{ secrets.LLM_BASE_URL }}
-          MAX_ITERATIONS: 15
-        run: |
-          echo "[llm.eval]" > config.toml
-          echo "model = \"$LLM_MODEL\"" >> config.toml
-          echo "api_key = \"$LLM_API_KEY\"" >> config.toml
-          echo "base_url = \"$LLM_BASE_URL\"" >> config.toml
-          echo "temperature = 0.0" >> config.toml
-      - name: Run integration test evaluation for VisualBrowsingAgent (DeepSeek)
-        env:
-          SANDBOX_FORCE_REBUILD_RUNTIME: True
-        run: |
-          poetry run ./evaluation/integration_tests/scripts/run_infer.sh llm.eval HEAD VisualBrowsingAgent '' 15 $N_PROCESSES "t05_simple_browsing,t06_github_pr_browsing.py" 'visualbrowsing_deepseek_run'
-
-          # Find and export the visual browsing agent test results
-          REPORT_FILE_VISUALBROWSING_DEEPSEEK=$(find evaluation/evaluation_outputs/outputs/integration_tests/VisualBrowsingAgent/deepseek*_maxiter_15_N* -name "report.md" -type f | head -n 1)
-          echo "REPORT_FILE_VISUALBROWSING_DEEPSEEK: $REPORT_FILE_VISUALBROWSING_DEEPSEEK"
-          echo "INTEGRATION_TEST_REPORT_VISUALBROWSING_DEEPSEEK<<EOF" >> $GITHUB_ENV
-          cat $REPORT_FILE_VISUALBROWSING_DEEPSEEK >> $GITHUB_ENV
-          echo >> $GITHUB_ENV
-          echo "EOF" >> $GITHUB_ENV
-
-      - name: Create archive of evaluation outputs
-        run: |
-          TIMESTAMP=$(date +'%y-%m-%d-%H-%M')
-          cd evaluation/evaluation_outputs/outputs  # Change to the outputs directory
-          tar -czvf ../../../integration_tests_${TIMESTAMP}.tar.gz integration_tests/CodeActAgent/* integration_tests/DelegatorAgent/* integration_tests/VisualBrowsingAgent/* # Only include the actual result directories
-
-      - name: Upload evaluation results as artifact
-        uses: actions/upload-artifact@v4
-        id: upload_results_artifact
-        with:
-          name: integration-test-outputs-${{ github.run_id }}-${{ github.run_attempt }}
-          path: integration_tests_*.tar.gz
-
-      - name: Get artifact URLs
-        run: |
-          echo "ARTIFACT_URL=${{ steps.upload_results_artifact.outputs.artifact-url }}" >> $GITHUB_ENV
-
-      - name: Set timestamp and trigger reason
-        run: |
-          echo "TIMESTAMP=$(date +'%Y-%m-%d-%H-%M')" >> $GITHUB_ENV
-          if [[ "${{ github.event_name }}" == "pull_request" ]]; then
-            echo "TRIGGER_REASON=pr-${{ github.event.pull_request.number }}" >> $GITHUB_ENV
-          elif [[ "${{ github.event_name }}" == "workflow_dispatch" ]]; then
-            echo "TRIGGER_REASON=manual-${{ github.event.inputs.reason }}" >> $GITHUB_ENV
-          else
-            echo "TRIGGER_REASON=nightly-scheduled" >> $GITHUB_ENV
-          fi
-
-      - name: Comment with results and artifact link
-        id: create_comment
-        uses: KeisukeYamashita/create-comment@v1
-        with:
-          # if triggered by PR, use PR number, otherwise use 5318 as fallback issue number for manual triggers
-          number: ${{ github.event_name == 'pull_request' && github.event.pull_request.number || 5318 }}
-          unique: false
-          comment: |
-              Trigger by: ${{ github.event_name == 'pull_request' && format('Pull Request (integration-test label on PR #{0})', github.event.pull_request.number) || (github.event_name == 'workflow_dispatch' && format('Manual Trigger: {0}', github.event.inputs.reason)) || 'Nightly Scheduled Run' }}
-              Commit: ${{ github.sha }}
-              **Integration Tests Report (Haiku)**
-              Haiku LLM Test Results:
-              ${{ env.INTEGRATION_TEST_REPORT_HAIKU }}
-              ---
-              **Integration Tests Report (DeepSeek)**
-              DeepSeek LLM Test Results:
-              ${{ env.INTEGRATION_TEST_REPORT_DEEPSEEK }}
-              ---
-                **Integration Tests Report Delegator (Haiku)**
-              ${{ env.INTEGRATION_TEST_REPORT_DELEGATOR_HAIKU }}
-              ---
-              **Integration Tests Report Delegator (DeepSeek)**
-              ${{ env.INTEGRATION_TEST_REPORT_DELEGATOR_DEEPSEEK }}
-              ---
-              **Integration Tests Report VisualBrowsing (DeepSeek)**
-              ${{ env.INTEGRATION_TEST_REPORT_VISUALBROWSING_DEEPSEEK }}
-              ---
-              Download testing outputs (includes both Haiku and DeepSeek results): [Download](${{ steps.upload_results_artifact.outputs.artifact-url }})

.github/workflows/lint-fix.yml

@@ -1,91 +0,0 @@
-name: Lint Fix
-
-on:
-  pull_request:
-    types: [labeled]
-
-jobs:
-  # Frontend lint fixes
-  lint-fix-frontend:
-    if: github.event.label.name == 'lint-fix'
-    name: Fix frontend linting issues
-    runs-on: ubuntu-latest
-    permissions:
-      contents: write
-      pull-requests: write
-    steps:
-      - uses: actions/checkout@v4
-        with:
-          ref: ${{ github.head_ref }}
-          repository: ${{ github.event.pull_request.head.repo.full_name }}
-          fetch-depth: 0
-          token: ${{ secrets.GITHUB_TOKEN }}
-
-      - name: Install Node.js 20
-        uses: actions/setup-node@v4
-        with:
-          node-version: 20
-      - name: Install frontend dependencies
-        run: |
-          cd frontend
-          npm install --frozen-lockfile
-      - name: Fix frontend lint issues
-        run: |
-          cd frontend
-          npm run lint:fix
-
-      # Commit and push changes if any
-      - name: Check for changes
-        id: git-check
-        run: |
-          git diff --quiet || echo "changes=true" >> $GITHUB_OUTPUT
-      - name: Commit and push if there are changes
-        if: steps.git-check.outputs.changes == 'true'
-        run: |
-          git config --local user.email "openhands@all-hands.dev"
-          git config --local user.name "OpenHands Bot"
-          git add -A
-          git commit -m "🤖 Auto-fix frontend linting issues"
-          git push
-
-  # Python lint fixes
-  lint-fix-python:
-    if: github.event.label.name == 'lint-fix'
-    name: Fix Python linting issues
-    runs-on: ubuntu-latest
-    permissions:
-      contents: write
-      pull-requests: write
-    steps:
-      - uses: actions/checkout@v4
-        with:
-          ref: ${{ github.head_ref }}
-          repository: ${{ github.event.pull_request.head.repo.full_name }}
-          fetch-depth: 0
-          token: ${{ secrets.GITHUB_TOKEN }}
-
-      - name: Set up python
-        uses: actions/setup-python@v5
-        with:
-          python-version: 3.12
-          cache: 'pip'
-      - name: Install pre-commit
-        run: pip install pre-commit==3.7.0
-      - name: Fix python lint issues
-        run: |
-          # Run all pre-commit hooks and continue even if they modify files (exit code 1)
-          pre-commit run --config ./dev_config/python/.pre-commit-config.yaml --files openhands/**/* evaluation/**/* tests/**/* || true
-
-      # Commit and push changes if any
-      - name: Check for changes
-        id: git-check
-        run: |
-          git diff --quiet || echo "changes=true" >> $GITHUB_OUTPUT
-      - name: Commit and push if there are changes
-        if: steps.git-check.outputs.changes == 'true'
-        run: |
-          git config --local user.email "openhands@all-hands.dev"
-          git config --local user.name "OpenHands Bot"
-          git add -A
-          git commit -m "🤖 Auto-fix Python linting issues"
-          git push

.github/workflows/lint.yml

@@ -1,68 +0,0 @@
-# Workflow that runs lint on the frontend and python code
-name: Lint
-
-# The jobs in this workflow are required, so they must run at all times
-# Always run on "main"
-# Always run on PRs
-on:
-  push:
-    branches:
-    - main
-  pull_request:
-
-# If triggered by a PR, it will be in the same group. However, each commit on main will be in its own unique group
-concurrency:
-  group: ${{ github.workflow }}-${{ (github.head_ref && github.ref) || github.run_id }}
-  cancel-in-progress: true
-
-jobs:
-  # Run lint on the frontend code
-  lint-frontend:
-    name: Lint frontend
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-      - name: Install Node.js 20
-        uses: actions/setup-node@v4
-        with:
-          node-version: 20
-      - name: Install dependencies
-        run: |
-          cd frontend
-          npm install --frozen-lockfile
-      - name: Lint and TypeScript compilation
-        run: |
-          cd frontend
-          npm run lint
-          npm run make-i18n && tsc
-
-  # Run lint on the python code
-  lint-python:
-    name: Lint python
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-        with:
-          fetch-depth: 0
-      - name: Set up python
-        uses: actions/setup-python@v5
-        with:
-          python-version: 3.12
-          cache: 'pip'
-      - name: Install pre-commit
-        run: pip install pre-commit==3.7.0
-      - name: Run pre-commit hooks
-        run: pre-commit run --files openhands/**/* evaluation/**/* tests/**/* --show-diff-on-failure --config ./dev_config/python/.pre-commit-config.yaml
-
-  # Check version consistency across documentation
-  check-version-consistency:
-    name: Check version consistency
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-      - name: Set up python
-        uses: actions/setup-python@v5
-        with:
-          python-version: 3.12
-      - name: Run version consistency check
-        run: .github/scripts/check_version_consistency.py

.github/workflows/openhands-resolver.yml

@@ -1,414 +0,0 @@
-name: Auto-Fix Tagged Issue with OpenHands
-
-on:
-  workflow_call:
-    inputs:
-      max_iterations:
-        required: false
-        type: number
-        default: 50
-      macro:
-        required: false
-        type: string
-        default: "@openhands-agent"
-      target_branch:
-        required: false
-        type: string
-        default: "main"
-        description: "Target branch to pull and create PR against"
-      LLM_MODEL:
-        required: false
-        type: string
-        default: "anthropic/claude-3-5-sonnet-20241022"
-      LLM_API_VERSION:
-        required: false
-        type: string
-        default: ""
-      base_container_image:
-        required: false
-        type: string
-        default: ""
-        description: "Custom sandbox env"
-    secrets:
-      LLM_MODEL:
-        required: false
-      LLM_API_KEY:
-        required: true
-      LLM_BASE_URL:
-        required: false
-      PAT_TOKEN:
-        required: false
-      PAT_USERNAME:
-        required: false
-
-  issues:
-    types: [labeled]
-  pull_request:
-    types: [labeled]
-  issue_comment:
-    types: [created]
-  pull_request_review_comment:
-    types: [created]
-  pull_request_review:
-    types: [submitted]
-
-permissions:
-  contents: write
-  pull-requests: write
-  issues: write
-
-jobs:
-  auto-fix:
-    if: |
-      github.event_name == 'workflow_call' ||
-      github.event.label.name == 'fix-me' ||
-      github.event.label.name == 'fix-me-experimental' ||
-      (
-        ((github.event_name == 'issue_comment' || github.event_name == 'pull_request_review_comment') &&
-        contains(github.event.comment.body, inputs.macro || '@openhands-agent') &&
-        (github.event.comment.author_association == 'OWNER' || github.event.comment.author_association == 'COLLABORATOR' || github.event.comment.author_association == 'MEMBER')
-        ) ||
-
-        (github.event_name == 'pull_request_review' &&
-        contains(github.event.review.body, inputs.macro || '@openhands-agent') &&
-        (github.event.review.author_association == 'OWNER' || github.event.review.author_association == 'COLLABORATOR' || github.event.review.author_association == 'MEMBER')
-        )
-      )
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout repository
-        uses: actions/checkout@v4
-
-      - name: Set up Python
-        uses: actions/setup-python@v5
-        with:
-          python-version: "3.12"
-
-      - name: Get latest versions and create requirements.txt
-        run: |
-          python -m pip index versions openhands-ai > openhands_versions.txt
-          OPENHANDS_VERSION=$(head -n 1 openhands_versions.txt | awk '{print $2}' | tr -d '()')
-
-          # Create a new requirements.txt locally within the workflow, ensuring no reference to the repo's file
-          echo "openhands-ai==${OPENHANDS_VERSION}" > /tmp/requirements.txt
-          cat /tmp/requirements.txt
-
-      - name: Cache pip dependencies
-        if: |
-          !(
-            github.event.label.name == 'fix-me-experimental' ||
-            (
-              (github.event_name == 'issue_comment' || github.event_name == 'pull_request_review_comment') &&
-              contains(github.event.comment.body, '@openhands-agent-exp')
-            ) ||
-            (
-              github.event_name == 'pull_request_review' &&
-              contains(github.event.review.body, '@openhands-agent-exp')
-            )
-          )
-        uses: actions/cache@v4
-        with:
-          path: ${{ env.pythonLocation }}/lib/python3.12/site-packages/*
-          key: ${{ runner.os }}-pip-openhands-resolver-${{ hashFiles('/tmp/requirements.txt') }}
-          restore-keys: |
-            ${{ runner.os }}-pip-openhands-resolver-${{ hashFiles('/tmp/requirements.txt') }}
-
-      - name: Check required environment variables
-        env:
-          LLM_MODEL: ${{ secrets.LLM_MODEL || inputs.LLM_MODEL }}
-          LLM_API_KEY: ${{ secrets.LLM_API_KEY }}
-          LLM_BASE_URL: ${{ secrets.LLM_BASE_URL }}
-          LLM_API_VERSION: ${{ inputs.LLM_API_VERSION }}
-          PAT_TOKEN: ${{ secrets.PAT_TOKEN }}
-          PAT_USERNAME: ${{ secrets.PAT_USERNAME }}
-          GITHUB_TOKEN: ${{ github.token }}
-        run: |
-          required_vars=("LLM_API_KEY")
-          for var in "${required_vars[@]}"; do
-            if [ -z "${!var}" ]; then
-              echo "Error: Required environment variable $var is not set."
-              exit 1
-            fi
-          done
-
-          # Check optional variables and warn about fallbacks
-          if [ -z "$LLM_BASE_URL" ]; then
-            echo "Warning: LLM_BASE_URL is not set, will use default API endpoint"
-          fi
-
-          if [ -z "$PAT_TOKEN" ]; then
-            echo "Warning: PAT_TOKEN is not set, falling back to GITHUB_TOKEN"
-          fi
-
-          if [ -z "$PAT_USERNAME" ]; then
-            echo "Warning: PAT_USERNAME is not set, will use openhands-agent"
-          fi
-
-      - name: Set environment variables
-        run: |
-          # Handle pull request events first
-          if [ -n "${{ github.event.pull_request.number }}" ]; then
-            echo "ISSUE_NUMBER=${{ github.event.pull_request.number }}" >> $GITHUB_ENV
-            echo "ISSUE_TYPE=pr" >> $GITHUB_ENV
-          # Handle pull request review events
-          elif [ -n "${{ github.event.review.body }}" ]; then
-            echo "ISSUE_NUMBER=${{ github.event.pull_request.number }}" >> $GITHUB_ENV
-            echo "ISSUE_TYPE=pr" >> $GITHUB_ENV
-          # Handle issue comment events that reference a PR
-          elif [ -n "${{ github.event.issue.pull_request }}" ]; then
-            echo "ISSUE_NUMBER=${{ github.event.issue.number }}" >> $GITHUB_ENV
-            echo "ISSUE_TYPE=pr" >> $GITHUB_ENV
-          # Handle regular issue events
-          else
-            echo "ISSUE_NUMBER=${{ github.event.issue.number }}" >> $GITHUB_ENV
-            echo "ISSUE_TYPE=issue" >> $GITHUB_ENV
-          fi
-
-          if [ -n "${{ github.event.review.body }}" ]; then
-            echo "COMMENT_ID=${{ github.event.review.id || 'None' }}" >> $GITHUB_ENV
-          else
-            echo "COMMENT_ID=${{ github.event.comment.id || 'None' }}" >> $GITHUB_ENV
-          fi
-
-          echo "MAX_ITERATIONS=${{ inputs.max_iterations || 50 }}" >> $GITHUB_ENV
-          echo "SANDBOX_ENV_GITHUB_TOKEN=${{ secrets.PAT_TOKEN || github.token }}" >> $GITHUB_ENV
-          echo "SANDBOX_ENV_BASE_CONTAINER_IMAGE=${{ inputs.base_container_image }}" >> $GITHUB_ENV
-
-          # Set branch variables
-          echo "TARGET_BRANCH=${{ inputs.target_branch || 'main' }}" >> $GITHUB_ENV
-
-      - name: Comment on issue with start message
-        uses: actions/github-script@v7
-        with:
-          github-token: ${{ secrets.PAT_TOKEN || github.token }}
-          script: |
-            const issueType = process.env.ISSUE_TYPE;
-            github.rest.issues.createComment({
-              issue_number: ${{ env.ISSUE_NUMBER }},
-              owner: context.repo.owner,
-              repo: context.repo.repo,
-              body: `[OpenHands](https://github.com/All-Hands-AI/OpenHands) started fixing the ${issueType}! You can monitor the progress [here](https://github.com/${context.repo.owner}/${context.repo.repo}/actions/runs/${context.runId}).`
-            });
-
-      - name: Install OpenHands
-        id: install_openhands
-        uses: actions/github-script@v7
-        env:
-          COMMENT_BODY: ${{ github.event.comment.body || '' }}
-          REVIEW_BODY: ${{ github.event.review.body || '' }}
-          LABEL_NAME: ${{ github.event.label.name || '' }}
-          EVENT_NAME: ${{ github.event_name }}
-        with:
-          script: |
-            const commentBody = process.env.COMMENT_BODY.trim();
-            const reviewBody = process.env.REVIEW_BODY.trim();
-            const labelName = process.env.LABEL_NAME.trim();
-            const eventName = process.env.EVENT_NAME.trim();
-            // Check conditions
-            const isExperimentalLabel = labelName === "fix-me-experimental";
-            const isIssueCommentExperimental =
-              (eventName === "issue_comment" || eventName === "pull_request_review_comment") &&
-              commentBody.includes("@openhands-agent-exp");
-            const isReviewCommentExperimental =
-              eventName === "pull_request_review" && reviewBody.includes("@openhands-agent-exp");
-
-            // Set output variable
-            core.setOutput('isExperimental', isExperimentalLabel || isIssueCommentExperimental || isReviewCommentExperimental);
-
-            // Perform package installation
-            if (isExperimentalLabel || isIssueCommentExperimental || isReviewCommentExperimental) {
-              console.log("Installing experimental OpenHands...");
-              await exec.exec("python -m pip install --upgrade pip");
-              await exec.exec("pip install git+https://github.com/all-hands-ai/openhands.git");
-            } else {
-              console.log("Installing from requirements.txt...");
-              await exec.exec("python -m pip install --upgrade pip");
-              await exec.exec("pip install -r /tmp/requirements.txt");
-            }
-
-      - name: Attempt to resolve issue
-        env:
-          GITHUB_TOKEN: ${{ secrets.PAT_TOKEN || github.token }}
-          GITHUB_USERNAME: ${{ secrets.PAT_USERNAME || 'openhands-agent' }}
-          GIT_USERNAME: ${{ secrets.PAT_USERNAME || 'openhands-agent' }}
-          LLM_MODEL: ${{ secrets.LLM_MODEL || inputs.LLM_MODEL }}
-          LLM_API_KEY: ${{ secrets.LLM_API_KEY }}
-          LLM_BASE_URL: ${{ secrets.LLM_BASE_URL }}
-          LLM_API_VERSION: ${{ inputs.LLM_API_VERSION }}
-          PYTHONPATH: ""
-        run: |
-          cd /tmp && python -m openhands.resolver.resolve_issue \
-            --repo ${{ github.repository }} \
-            --issue-number ${{ env.ISSUE_NUMBER }} \
-            --issue-type ${{ env.ISSUE_TYPE }} \
-            --max-iterations ${{ env.MAX_ITERATIONS }} \
-            --comment-id ${{ env.COMMENT_ID }} \
-            --is-experimental ${{ steps.install_openhands.outputs.isExperimental }}
-
-      - name: Check resolution result
-        id: check_result
-        run: |
-          if cd /tmp && grep -q '"success":true' output/output.jsonl; then
-            echo "RESOLUTION_SUCCESS=true" >> $GITHUB_OUTPUT
-          else
-            echo "RESOLUTION_SUCCESS=false" >> $GITHUB_OUTPUT
-          fi
-
-      - name: Upload output.jsonl as artifact
-        uses: actions/upload-artifact@v4
-        if: always() # Upload even if the previous steps fail
-        with:
-          name: resolver-output
-          path: /tmp/output/output.jsonl
-          retention-days: 30 # Keep the artifact for 30 days
-
-      - name: Create draft PR or push branch
-        if: always() # Create PR or branch even if the previous steps fail
-        env:
-          GITHUB_TOKEN: ${{ secrets.PAT_TOKEN || github.token }}
-          GITHUB_USERNAME: ${{ secrets.PAT_USERNAME || 'openhands-agent' }}
-          GIT_USERNAME: ${{ secrets.PAT_USERNAME || 'openhands-agent' }}
-          LLM_MODEL: ${{ secrets.LLM_MODEL || inputs.LLM_MODEL }}
-          LLM_API_KEY: ${{ secrets.LLM_API_KEY }}
-          LLM_BASE_URL: ${{ secrets.LLM_BASE_URL }}
-          LLM_API_VERSION: ${{ inputs.LLM_API_VERSION }}
-          PYTHONPATH: ""
-        run: |
-          if [ "${{ steps.check_result.outputs.RESOLUTION_SUCCESS }}" == "true" ]; then
-            cd /tmp && python -m openhands.resolver.send_pull_request \
-              --issue-number ${{ env.ISSUE_NUMBER }} \
-              --target-branch ${{ env.TARGET_BRANCH }} \
-              --pr-type draft \
-              --reviewer ${{ github.actor }} | tee pr_result.txt && \
-              grep "draft created" pr_result.txt | sed 's/.*\///g' > pr_number.txt
-          else
-            cd /tmp && python -m openhands.resolver.send_pull_request \
-              --issue-number ${{ env.ISSUE_NUMBER }} \
-              --pr-type branch \
-              --send-on-failure | tee branch_result.txt && \
-              grep "branch created" branch_result.txt | sed 's/.*\///g; s/.expand=1//g' > branch_name.txt
-          fi
-
-      # Step leaves comment for when agent is invoked on PR
-      - name: Analyze Push Logs (Updated PR or No Changes) # Skip comment if PR update was successful OR leave comment if the agent made no code changes
-        uses: actions/github-script@v7
-        if: always()
-        env:
-          AGENT_RESPONDED: ${{ env.AGENT_RESPONDED || 'false' }}
-        with:
-          github-token: ${{ secrets.PAT_TOKEN || github.token }}
-          script: |
-            const fs = require('fs');
-            const issueNumber = ${{ env.ISSUE_NUMBER }};
-            let logContent = '';
-
-            try {
-              logContent = fs.readFileSync('/tmp/pr_result.txt', 'utf8').trim();
-            } catch (error) {
-              console.error('Error reading pr_result.txt file:', error);
-            }
-
-            const noChangesMessage = `No changes to commit for issue #${issueNumber}. Skipping commit.`;
-
-            // Check logs from send_pull_request.py (pushes code to GitHub)
-            if (logContent.includes("Updated pull request")) {
-              console.log("Updated pull request found. Skipping comment.");
-              process.env.AGENT_RESPONDED = 'true';
-            } else if (logContent.includes(noChangesMessage)) {
-              github.rest.issues.createComment({
-                issue_number: issueNumber,
-                owner: context.repo.owner,
-                repo: context.repo.repo,
-                body: `The workflow to fix this issue encountered an error. Openhands failed to create any code changes.`
-              });
-              process.env.AGENT_RESPONDED = 'true';
-            }
-
-      # Step leaves comment for when agent is invoked on issue
-      - name: Comment on issue # Comment link to either PR or branch created by agent
-        uses: actions/github-script@v7
-        if: always() # Comment on issue even if the previous steps fail
-        env:
-          AGENT_RESPONDED: ${{ env.AGENT_RESPONDED || 'false' }}
-        with:
-          github-token: ${{ secrets.PAT_TOKEN || github.token }}
-          script: |
-            const fs = require('fs');
-            const path = require('path');
-            const issueNumber = ${{ env.ISSUE_NUMBER }};
-            const success = ${{ steps.check_result.outputs.RESOLUTION_SUCCESS }};
-
-            let prNumber = '';
-            let branchName = '';
-            let resultExplanation = '';
-
-            try {
-              if (success) {
-                prNumber = fs.readFileSync('/tmp/pr_number.txt', 'utf8').trim();
-              } else {
-                branchName = fs.readFileSync('/tmp/branch_name.txt', 'utf8').trim();
-              }
-            } catch (error) {
-              console.error('Error reading file:', error);
-            }
-
-
-            try {
-              if (!success){
-                // Read result_explanation from JSON file for failed resolution
-                const outputFilePath = path.resolve('/tmp/output/output.jsonl');
-                if (fs.existsSync(outputFilePath)) {
-                  const outputContent = fs.readFileSync(outputFilePath, 'utf8');
-                  const jsonLines = outputContent.split('\n').filter(line => line.trim() !== '');
-
-                  if (jsonLines.length > 0) {
-                    // First entry in JSON lines has the key 'result_explanation'
-                    const firstEntry = JSON.parse(jsonLines[0]);
-                    resultExplanation = firstEntry.result_explanation || '';
-                  }
-                }
-              }
-            } catch (error){
-              console.error('Error reading file:', error);
-            }
-
-            // Check "success" log from resolver output
-            if (success && prNumber) {
-              github.rest.issues.createComment({
-                issue_number: issueNumber,
-                owner: context.repo.owner,
-                repo: context.repo.repo,
-                body: `A potential fix has been generated and a draft PR #${prNumber} has been created. Please review the changes.`
-              });
-              process.env.AGENT_RESPONDED = 'true';
-            } else if (!success && branchName) {
-              let commentBody = `An attempt was made to automatically fix this issue, but it was unsuccessful. A branch named '${branchName}' has been created with the attempted changes. You can view the branch [here](https://github.com/${context.repo.owner}/${context.repo.repo}/tree/${branchName}). Manual intervention may be required.`;
-
-              if (resultExplanation) {
-                commentBody += `\n\nAdditional details about the failure:\n${resultExplanation}`;
-              }
-
-              github.rest.issues.createComment({
-                issue_number: issueNumber,
-                owner: context.repo.owner,
-                repo: context.repo.repo,
-                body: commentBody
-              });
-              process.env.AGENT_RESPONDED = 'true';
-            }
-
-      # Leave error comment when both PR/Issue comment handling fail
-      - name: Fallback Error Comment
-        uses: actions/github-script@v7
-        if: ${{ env.AGENT_RESPONDED == 'false' }} # Only run if no conditions were met in previous steps
-        with:
-          github-token: ${{ secrets.PAT_TOKEN || github.token }}
-          script: |
-            const issueNumber = ${{ env.ISSUE_NUMBER }};
-
-            github.rest.issues.createComment({
-              issue_number: issueNumber,
-              owner: context.repo.owner,
-              repo: context.repo.repo,
-              body: `The workflow to fix this issue encountered an error. Please check the [workflow logs](https://github.com/${context.repo.owner}/${context.repo.repo}/actions/runs/${context.runId}) for more information.`
-            });

.github/workflows/py-unit-tests.yml

@@ -1,55 +0,0 @@
-# Workflow that runs python unit tests
-name: Run Python Unit Tests
-
-# The jobs in this workflow are required, so they must run at all times
-# * Always run on "main"
-# * Always run on PRs
-on:
-  push:
-    branches:
-      - main
-  pull_request:
-
-# If triggered by a PR, it will be in the same group. However, each commit on main will be in its own unique group
-concurrency:
-  group: ${{ github.workflow }}-${{ (github.head_ref && github.ref) || github.run_id }}
-  cancel-in-progress: true
-
-jobs:
-  # Run python unit tests on Linux
-  test-on-linux:
-    name: Python Unit Tests on Linux
-    runs-on: ubuntu-latest
-    env:
-      INSTALL_DOCKER: '0' # Set to '0' to skip Docker installation
-    strategy:
-      matrix:
-        python-version: ['3.12']
-    steps:
-      - uses: actions/checkout@v4
-      - name: Set up Docker Buildx
-        id: buildx
-        uses: docker/setup-buildx-action@v3
-      - name: Install tmux
-        run: sudo apt-get update && sudo apt-get install -y tmux
-      - name: Setup Node.js
-        uses: actions/setup-node@v4
-        with:
-          node-version: '22.x'
-      - name: Install poetry via pipx
-        run: pipx install poetry
-      - name: Set up Python
-        uses: actions/setup-python@v5
-        with:
-          python-version: ${{ matrix.python-version }}
-          cache: 'poetry'
-      - name: Install Python dependencies using Poetry
-        run: poetry install --without evaluation,llama-index
-      - name: Build Environment
-        run: make build
-      - name: Run Tests
-        run: poetry run pytest --forked -n auto --cov=openhands --cov-report=xml -svv ./tests/unit --ignore=tests/unit/test_long_term_memory.py
-      - name: Upload coverage to Codecov
-        uses: codecov/codecov-action@v5
-        env:
-          CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}

.github/workflows/pypi-release.yml

@@ -1,31 +0,0 @@
-# Publishes the OpenHands PyPi package
-name: Publish PyPi Package
-
-# Triggered manually
-on:
-  workflow_dispatch:
-    inputs:
-      reason:
-        description: 'Reason for manual trigger'
-        required: true
-        default: ''
-
-jobs:
-  release:
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-      - uses: actions/setup-python@v5
-        with:
-          python-version: 3.12
-      - name: Install Poetry
-        uses: snok/install-poetry@v1.4.1
-        with:
-          virtualenvs-in-project: true
-          virtualenvs-path: ~/.virtualenvs
-      - name: Install Poetry Dependencies
-        run: poetry install --no-interaction --no-root
-      - name: Build poetry project
-        run: ./build.sh
-      - name: publish
-        run: poetry publish -u __token__ -p ${{ secrets.PYPI_TOKEN }}

.github/workflows/run-eval.yml

@@ -1,53 +0,0 @@
-# Run evaluation on a PR
-name: Run Eval
-
-# Runs when a PR is labeled with one of the "run-eval-" labels
-on:
-  pull_request:
-    types: [labeled]
-
-jobs:
-  trigger-job:
-    name: Trigger remote eval job
-    if: ${{ github.event.label.name == 'run-eval-xs' || github.event.label.name == 'run-eval-s' || github.event.label.name == 'run-eval-m' }}
-    runs-on: ubuntu-latest
-
-    steps:
-      - name: Checkout PR branch
-        uses: actions/checkout@v4
-        with:
-          ref: ${{ github.head_ref }}
-
-      - name: Trigger remote job
-        run: |
-          REPO_URL="https://github.com/${{ github.repository }}"
-          PR_BRANCH="${{ github.head_ref }}"
-          echo "Repository URL: $REPO_URL"
-          echo "PR Branch: $PR_BRANCH"
-
-          if [[ "${{ github.event.label.name }}" == "run-eval-xs" ]]; then
-            EVAL_INSTANCES="1"
-          elif [[ "${{ github.event.label.name }}" == "run-eval-s" ]]; then
-            EVAL_INSTANCES="5"
-          elif [[ "${{ github.event.label.name }}" == "run-eval-m" ]]; then
-            EVAL_INSTANCES="30"
-          fi
-
-          curl -X POST \
-            -H "Authorization: Bearer ${{ secrets.PAT_TOKEN }}" \
-            -H "Accept: application/vnd.github+json" \
-            -d "{\"ref\": \"main\", \"inputs\": {\"github-repo\": \"${REPO_URL}\", \"github-branch\": \"${PR_BRANCH}\", \"pr-number\": \"${{ github.event.pull_request.number }}\", \"eval-instances\": \"${EVAL_INSTANCES}\"}}" \
-            https://api.github.com/repos/All-Hands-AI/evaluation/actions/workflows/create-branch.yml/dispatches
-
-          # Send Slack message
-          PR_URL="https://github.com/${{ github.repository }}/pull/${{ github.event.pull_request.number }}"
-          slack_text="PR $PR_URL has triggered evaluation on $EVAL_INSTANCES instances..."
-          curl -X POST -H 'Content-type: application/json' --data '{"text":"'"$slack_text"'"}' \
-            https://hooks.slack.com/services/${{ secrets.SLACK_TOKEN }}
-
-      - name: Comment on PR
-        uses: KeisukeYamashita/create-comment@v1
-        with:
-          unique: false
-          comment: |
-            Running evaluation on the PR. Once eval is done, the results will be posted.

.github/workflows/stale.yml

@@ -1,22 +0,0 @@
-# Workflow that marks issues and PRs with no activity for 30 days with "Stale" and closes them after 7 more days of no activity
-name: 'Close stale issues'
-
-# Runs every day at 01:30
-on:
-  schedule:
-    - cron: '30 1 * * *'
-
-jobs:
-  stale:
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/stale@v9
-        with:
-          stale-issue-message: 'This issue is stale because it has been open for 30 days with no activity. Remove stale label or comment or this will be closed in 7 days.'
-          stale-pr-message: 'This PR is stale because it has been open for 30 days with no activity. Remove stale label or comment or this will be closed in 7 days.'
-          days-before-stale: 30
-          exempt-issue-labels: 'tracked'
-          close-issue-message: 'This issue was closed because it has been stalled for over 30 days with no activity.'
-          close-pr-message: 'This PR was closed because it has been stalled for over 30 days with no activity.'
-          days-before-close: 7
-          operations-per-run: 150

.gitignore

@@ -1,235 +0,0 @@
-# Byte-compiled / optimized / DLL files
-__pycache__/
-*.py[cod]
-*$py.class
-
-# C extensions
-*.so
-
-# Distribution / packaging
-.Python
-build/
-develop-eggs/
-dist/
-downloads/
-eggs/
-.eggs/
-./lib/
-lib64/
-parts/
-sdist/
-var/
-wheels/
-share/python-wheels/
-*.egg-info/
-.installed.cfg
-*.egg
-MANIFEST
-requirements.txt
-
-# PyInstaller
-#  Usually these files are written by a python script from a template
-#  before PyInstaller builds the exe, so as to inject date/other infos into it.
-*.manifest
-*.spec
-
-# Installer logs
-pip-log.txt
-pip-delete-this-directory.txt
-
-# Unit test / coverage reports
-htmlcov/
-.tox/
-.nox/
-.coverage
-.coverage.*
-.cache
-nosetests.xml
-coverage.xml
-*.cover
-*.py,cover
-.hypothesis/
-.pytest_cache/
-cover/
-
-# Translations
-*.mo
-*.pot
-
-# Django stuff:
-local_settings.py
-db.sqlite3
-db.sqlite3-journal
-
-# Flask stuff:
-instance/
-.webassets-cache
-
-# Scrapy stuff:
-.scrapy
-
-# Sphinx documentation
-docs/_build/
-
-# PyBuilder
-.pybuilder/
-target/
-
-# Jupyter Notebook
-.ipynb_checkpoints
-
-# IPython
-profile_default/
-ipython_config.py
-
-# pyenv
-#   For a library or package, you might want to ignore these files since the code is
-#   intended to run in multiple environments; otherwise, check them in:
-.python-version
-
-# pipenv
-#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
-#   However, in case of collaboration, if having platform-specific dependencies or dependencies
-#   having no cross-platform support, pipenv may install dependencies that don't work, or not
-#   install all needed dependencies.
-#Pipfile.lock
-
-# poetry
-#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
-#   This is especially recommended for binary packages to ensure reproducibility, and is more
-#   commonly ignored for libraries.
-#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
-# poetry.lock
-
-# pdm
-#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
-#pdm.lock
-#   pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
-#   in version control.
-#   https://pdm.fming.dev/#use-with-ide
-.pdm.toml
-
-# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
-__pypackages__/
-
-# Celery stuff
-celerybeat-schedule
-celerybeat.pid
-
-# SageMath parsed files
-*.sage.py
-
-# Environments
-.env
-frontend/.env
-.venv
-env/
-venv/
-ENV/
-env.bak/
-.env.bak
-venv.bak/
-*venv/
-
-# Spyder project settings
-.spyderproject
-.spyproject
-
-# Rope project settings
-.ropeproject
-
-# mkdocs documentation
-/site
-
-# mypy
-.mypy_cache/
-.dmypy.json
-dmypy.json
-
-# Pyre type checker
-.pyre/
-
-# pytype static type analyzer
-.pytype/
-
-# Cython debug symbols
-cython_debug/
-
-# PyCharm
-#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
-#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
-#  and can be added to the global gitignore or merged into this file.  For a more nuclear
-#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
-.idea/
-.vscode/
-.cursorignore
-
-# evaluation
-evaluation/evaluation_outputs
-evaluation/outputs
-evaluation/swe_bench/eval_workspace*
-evaluation/SWE-bench/data
-evaluation/webarena/scripts/webarena_env.sh
-evaluation/bird/data
-evaluation/gaia/data
-evaluation/gorilla/data
-evaluation/toolqa/data
-evaluation/scienceagentbench/benchmark
-evaluation/commit0_bench/repos
-
-# openhands resolver
-output/
-
-# frontend
-
-# dependencies
-frontend/.pnp
-frontend/bun.lockb
-frontend/yarn.lock
-.pnp.js
-
-# testing
-frontend/coverage
-test_results*
-/_test_files_tmp/
-
-# production
-frontend/build
-frontend/dist
-
-# misc
-.DS_Store
-.env.local
-.env.development.local
-.env.test.local
-.env.production.local
-
-npm-debug.log*
-yarn-debug.log*
-yarn-error.log*
-
-logs
-
-# agent
-.envrc
-/workspace
-/_test_workspace
-/debug
-cache
-
-# configuration
-config.toml
-config.toml_
-config.toml.bak
-
-# swe-bench-eval
-image_build_logs
-run_instance_logs
-
-runtime_*.tar
-
-# docker build
-containers/runtime/Dockerfile
-containers/runtime/project.tar.gz
-containers/runtime/code
-**/node_modules/

.nvmrc

@@ -1 +0,0 @@
-22

.openhands/microagents/glossary.md

@@ -1,172 +0,0 @@
-# OpenHands Glossary
-
-### Agent
-The core AI entity in OpenHands that can perform software development tasks by interacting with tools, browsing the web, and modifying code.
-
-#### Agent Controller
-A component that manages the agent's lifecycle, handles its state, and coordinates interactions between the agent and various tools.
-
-#### Agent Delegation
-The ability of an agent to hand off specific tasks to other specialized agents for better task completion.
-
-#### Agent Hub
-A central registry of different agent types and their capabilities, allowing for easy agent selection and instantiation.
-
-#### Agent Skill
-A specific capability or function that an agent can perform, such as file manipulation, web browsing, or code editing.
-
-#### Agent State
-The current context and status of an agent, including its memory, active tools, and ongoing tasks.
-
-#### CodeAct Agent
-[A generalist agent in OpenHands](https://arxiv.org/abs/2407.16741) designed to perform tasks by editing and executing code.
-
-### Browser
-A system for web-based interactions and tasks.
-
-#### Browser Gym
-A testing and evaluation environment for browser-based agent interactions and tasks.
-
-#### Web Browser Tool
-A tool that enables agents to interact with web pages and perform web-based tasks.
-
-### Commands
-Terminal and execution related functionality.
-
-#### Bash Session
-A persistent terminal session that maintains state and history for bash command execution.
-This uses tmux under the hood.
-
-### Configuration
-System-wide settings and options.
-
-#### Agent Configuration
-Settings that define an agent's behavior, capabilities, and limitations, including available tools and runtime settings.
-
-#### Configuration Options
-Settings that control various aspects of OpenHands behavior, including runtime, security, and agent settings.
-
-#### LLM Config
-Configuration settings for language models used by agents, including model selection and parameters.
-
-#### LLM Draft Config
-Settings for draft mode operations with language models, typically used for faster, lower-quality responses.
-
-#### Runtime Configuration
-Settings that define how the runtime environment should be set up and operated.
-
-#### Security Options
-Configuration settings that control security features and restrictions.
-
-### Conversation
-A sequence of interactions between a user and an agent, including messages, actions, and their results.
-
-#### Conversation Info
-Metadata about a conversation, including its status, participants, and timeline.
-
-#### Conversation Manager
-A component that handles the creation, storage, and retrieval of conversations.
-
-#### Conversation Metadata
-Additional information about conversations, such as tags, timestamps, and related resources.
-
-#### Conversation Status
-The current state of a conversation, including whether it's active, completed, or failed.
-
-#### Conversation Store
-A storage system for maintaining conversation history and related data.
-
-### Events
-
-#### Event
-Every Conversation comprises a series of Events. Each Event is either an Action or an Observation.
-
-#### Event Stream
-A continuous flow of events that represents the ongoing activities and interactions in the system.
-
-#### Action
-A specific operation or command that an agent executes through available tools, such as running a command or editing a file.
-
-#### Observation
-The response or result returned by a tool after an agent's action, providing feedback about the action's outcome.
-
-### Interface
-Different ways to interact with OpenHands.
-
-#### CLI Mode
-A command-line interface mode for interacting with OpenHands agents without a graphical interface.
-
-#### GUI Mode
-A graphical user interface mode for interacting with OpenHands agents through a web interface.
-
-#### Headless Mode
-A mode of operation where OpenHands runs without a user interface, suitable for automation and scripting.
-
-### Agent Memory
-The system that decides which parts of the Event Stream (i.e. the conversation history) should be passed into each LLM prompt.
-
-#### Memory Store
-A storage system for maintaining agent memory and context across sessions.
-
-#### Condenser
-A component that processes and summarizes conversation history to maintain context while staying within token limits.
-
-#### Truncation
-A very simple Condenser strategy. Reduces conversation history or content to stay within token limits.
-
-### Microagent
-A specialized prompt that enhances OpenHands with domain-specific knowledge, repository-specific context, and task-specific workflows.
-
-#### Microagent Registry
-A central repository of available microagents and their configurations.
-
-#### Public Microagent
-A general-purpose microagent available to all OpenHands users, triggered by specific keywords.
-
-#### Repository Microagent
-A type of microagent that provides repository-specific context and guidelines, stored in the `.openhands/microagents/` directory.
-
-### Prompt
-Components for managing and processing prompts.
-
-#### Prompt Caching
-A system for caching and reusing common prompts to improve performance.
-
-#### Prompt Manager
-A component that handles the loading, processing, and management of prompts used by agents, including microagents.
-
-#### Response Parsing
-The process of interpreting and structuring responses from language models and tools.
-
-### Runtime
-The execution environment where agents perform their tasks, which can be local, remote, or containerized.
-
-#### Action Execution Server
-A REST API that receives agent actions (e.g. bash commands, python code, browsing actions), executes them in the runtime environment, and returns the results.
-
-#### Action Execution Client
-A component that handles the execution of actions in the runtime environment, managing the communication between the agent and the runtime.
-
-#### Docker Runtime
-A containerized runtime environment that provides isolation and reproducibility for agent operations.
-
-#### E2B Runtime
-A specialized runtime environment built on E2B for secure and isolated code execution.
-
-#### Local Runtime
-A runtime environment that executes on the local machine, suitable for development and testing.
-
-#### Modal Runtime
-A runtime environment built on Modal for scalable and distributed agent operations.
-
-#### Remote Runtime
-A sandboxed environment that executes code and commands remotely, providing isolation and security for agent operations.
-
-#### Runtime Builder
-A component that builds a Docker image for the Action Execution Server based on a user-specified base image.
-
-### Security
-Security-related components and features.
-
-#### Security Analyzer
-A component that checks agent actions for potential security risks.

.openhands/microagents/repo.md

@@ -1,42 +0,0 @@
----
-name: repo
-type: repo
-agent: CodeActAgent
----
-This repository contains the code for OpenHands, an automated AI software engineer. It has a Python backend
-(in the `openhands` directory) and React frontend (in the `frontend` directory).
-
-## General Setup:
-To set up the entire repo, including frontend and backend, run `make build`.
-You don't need to do this unless the user asks you to, or if you're trying to run the entire application.
-
-Before pushing any changes, you should ensure that any lint errors or simple test errors have been fixed.
-
-* If you've made changes to the backend, you should run `pre-commit run --all-files --config ./dev_config/python/.pre-commit-config.yaml`
-* If you've made changes to the frontend, you should run `cd frontend && npm run lint:fix && npm run build ; cd ..`
-
-If either command fails, it may have automatically fixed some issues. You should fix any issues that weren't automatically fixed,
-then re-run the command to ensure it passes.
-
-## Repository Structure
-Backend:
-- Located in the `openhands` directory
-- Testing:
-  - All tests are in `tests/unit/test_*.py`
-  - To test new code, run `poetry run pytest tests/unit/test_xxx.py` where `xxx` is the appropriate file for the current functionality
-  - Write all tests with pytest
-
-Frontend:
-- Located in the `frontend` directory
-- Prerequisites: A recent version of NodeJS / NPM
-- Setup: Run `npm install` in the frontend directory
-- Testing:
-  - Run tests: `npm run test`
-  - To run specific tests: `npm run test -- -t "TestName"`
-- Building:
-  - Build for production: `npm run build`
-- Environment Variables:
-  - Set in `frontend/.env` or as environment variables
-  - Available variables: VITE_BACKEND_HOST, VITE_USE_TLS, VITE_INSECURE_SKIP_VERIFY, VITE_FRONTEND_PORT
-- Internationalization:
-  - Generate i18n declaration file: `npm run make-i18n`

CITATION.cff

@@ -1,55 +0,0 @@
-cff-version: 1.2.0
-message: "If you use this software, please cite it using the following metadata."
-title: "OpenHands: An Open Platform for AI Software Developers as Generalist Agents"
-authors:
-  - family-names: Wang
-    given-names: Xingyao
-  - family-names: Li
-    given-names: Boxuan
-  - family-names: Song
-    given-names: Yufan
-  - family-names: Xu
-    given-names: Frank F.
-  - family-names: Tang
-    given-names: Xiangru
-  - family-names: Zhuge
-    given-names: Mingchen
-  - family-names: Pan
-    given-names: Jiayi
-  - family-names: Song
-    given-names: Yueqi
-  - family-names: Li
-    given-names: Bowen
-  - family-names: Singh
-    given-names: Jaskirat
-  - family-names: Tran
-    given-names: Hoang H.
-  - family-names: Li
-    given-names: Fuqiang
-  - family-names: Ma
-    given-names: Ren
-  - family-names: Zheng
-    given-names: Mingzhang
-  - family-names: Qian
-    given-names: Bill
-  - family-names: Shao
-    given-names: Yanjun
-  - family-names: Muennighoff
-    given-names: Niklas
-  - family-names: Zhang
-    given-names: Yizhe
-  - family-names: Hui
-    given-names: Binyuan
-  - family-names: Lin
-    given-names: Junyang
-  - family-names: Brennan
-    given-names: Robert
-  - family-names: Peng
-    given-names: Hao
-  - family-names: Ji
-    given-names: Heng
-  - family-names: Neubig
-    given-names: Graham
-year: 2024
-doi: "10.48550/arXiv.2407.16741"
-url: "https://arxiv.org/abs/2407.16741"

CODE_OF_CONDUCT.md

@@ -1,147 +0,0 @@
-
-# Contributor Covenant Code of Conduct
-
-## Our Pledge
-
-We as members, contributors, and leaders pledge to make participation in our
-community a harassment-free experience for everyone, regardless of age, body
-size, visible or invisible disability, ethnicity, sex characteristics, gender
-identity and expression, level of experience, education, socio-economic status,
-nationality, personal appearance, race, caste, color, religion, or sexual
-identity and orientation.
-
-We pledge to act and interact in ways that contribute to an open, welcoming,
-diverse, inclusive, and healthy community.
-
-## Our Standards
-
-Examples of behavior that contributes to a positive environment for our
-community include:
-
-* Demonstrating empathy and kindness toward other people.
-* Being respectful of differing opinions, viewpoints, and experiences.
-* Giving and gracefully accepting constructive feedback.
-* Accepting responsibility and apologizing to those affected by our mistakes,
-  and learning from the experience.
-* Focusing on what is best not just for us as individuals, but for the overall
-  community.
-
-Examples of unacceptable behavior include:
-
-* The use of sexualized language or imagery, and sexual attention or advances of
-  any kind.
-* Trolling, insulting or derogatory comments, and personal or political attacks.
-* Public or private harassment.
-* Publishing others' private information, such as a physical or email address,
-  without their explicit permission.
-* Other conduct which could reasonably be considered inappropriate in a
-  professional setting.
-
-## Enforcement Responsibilities
-
-Community leaders are responsible for clarifying and enforcing our standards of
-acceptable behavior and will take appropriate and fair corrective action in
-response to any behavior that they deem inappropriate, threatening, offensive,
-or harmful.
-
-Community leaders have the right and responsibility to remove, edit, or reject
-comments, commits, code, wiki edits, issues, and other contributions that are
-not aligned to this Code of Conduct, and will communicate reasons for moderation
-decisions when appropriate.
-
-## Scope
-
-This Code of Conduct applies within all community spaces, and also applies when
-an individual is officially representing the community in public spaces.
-Examples of representing our community include using an official email address,
-posting via an official social media account, or acting as an appointed
-representative at an online or offline event.
-
-## Enforcement
-
-Instances of abusive, harassing, or otherwise unacceptable behavior may be
-reported to the community leaders responsible for enforcement at
-contact@all-hands.dev.
-All complaints will be reviewed and investigated promptly and fairly.
-
-All community leaders are obligated to respect the privacy and security of the
-reporter of any incident.
-
-## Enforcement Guidelines
-
-Community leaders will follow these Community Impact Guidelines in determining
-the consequences for any action they deem in violation of this Code of Conduct:
-
-### 1. Correction
-
-**Community Impact**: Use of inappropriate language or other behavior deemed
-unprofessional or unwelcome in the community.
-
-**Consequence**: A private, written warning from community leaders, providing
-clarity around the nature of the violation and an explanation of why the
-behavior was inappropriate. A public apology may be requested.
-
-### 2. Warning
-
-**Community Impact**: A violation through a single incident or series of
-actions.
-
-**Consequence**: A warning with consequences for continued behavior. No
-interaction with the people involved, including unsolicited interaction with
-those enforcing the Code of Conduct, for a specified period of time. This
-includes avoiding interactions in community spaces as well as external channels
-like social media. Violating these terms may lead to a temporary or permanent
-ban.
-
-### 3. Temporary Ban
-
-**Community Impact**: A serious violation of community standards, including
-sustained inappropriate behavior.
-
-**Consequence**: A temporary ban from any sort of interaction or public
-communication with the community for a specified period of time. No public or
-private interaction with the people involved, including unsolicited interaction
-with those enforcing the Code of Conduct, is allowed during this period.
-Violating these terms may lead to a permanent ban.
-
-### 4. Permanent Ban
-
-**Community Impact**: Demonstrating a pattern of violation of community
-standards, including sustained inappropriate behavior, harassment of an
-individual, or aggression toward or disparagement of classes of individuals.
-
-**Consequence**: A permanent ban from any sort of public interaction within the
-community.
-
-### Slack and Discord Etiquettes
-
-These Slack and Discord etiquette guidelines are designed to foster an inclusive, respectful, and productive environment for all community members. By following these best practices, we ensure effective communication and collaboration while minimizing disruptions. Let’s work together to build a supportive and welcoming community!
-
-- Communicate respectfully and professionally, avoiding sarcasm or harsh language, and remember that tone can be difficult to interpret in text.
-- Use threads for specific discussions to keep channels organized and easier to follow.
-- Tag others only when their input is critical or urgent, and use @here, @channel or @everyone sparingly to minimize disruptions.
-- Be patient, as open-source contributors and maintainers often have other commitments and may need time to respond.
-- Post questions or discussions in the most relevant channel (e.g., for [slack - #general](https://app.slack.com/client/T06P212QSEA/C06P5NCGSFP) for general topics, [slack - #questions](https://openhands-ai.slack.com/archives/C06U8UTKSAD) for queries/questions, [discord - #general](https://discord.com/channels/1222935860639563850/1222935861386018885)).
-- When asking for help or raising issues, include necessary details like links, screenshots, or clear explanations to provide context.
-- Keep discussions in public channels whenever possible to allow others to benefit from the conversation, unless the matter is sensitive or private.
-- Always adhere to [our standards](https://github.com/All-Hands-AI/OpenHands/blob/main/CODE_OF_CONDUCT.md#our-standards) to ensure a welcoming and collaborative environment.
-- If you choose to mute a channel, consider setting up alerts for topics that still interest you to stay engaged. For Slack, Go to Settings → Notifications → My Keywords to add specific keywords that will notify you when mentioned. For example, if you're here for discussions about LLMs, mute the channel if it’s too busy, but set notifications to alert you only when “LLMs” appears in messages. Also for Discord, go to the channel notifications and choose the option that best describes your need.
-
-## Attribution
-
-This Code of Conduct is adapted from the [Contributor Covenant][homepage],
-version 2.1, available at
-[https://www.contributor-covenant.org/version/2/1/code_of_conduct.html][v2.1].
-
-Community Impact Guidelines were inspired by
-[Mozilla's code of conduct enforcement ladder][Mozilla CoC].
-
-For answers to common questions about this code of conduct, see the FAQ at
-[https://www.contributor-covenant.org/faq][FAQ]. Translations are available at
-[https://www.contributor-covenant.org/translations][translations].
-
-[homepage]: https://www.contributor-covenant.org
-[v2.1]: https://www.contributor-covenant.org/version/2/1/code_of_conduct.html
-[Mozilla CoC]: https://github.com/mozilla/diversity
-[FAQ]: https://www.contributor-covenant.org/faq
-[translations]: https://www.contributor-covenant.org/translations

COMMUNITY.md

@@ -1,43 +0,0 @@
-# 🙌 The OpenHands Community
-
-The OpenHands community is built around the belief that (1) AI and AI agents are going to fundamentally change the way
-we build software, and (2) if this is true, we should do everything we can to make sure that the benefits provided by
-such powerful technology are accessible to everyone.
-
-If this resonates with you, we'd love to have you join us in our quest!
-
-## 🤝 How to Join
-
-Check out our [How to Join the Community section.](https://github.com/All-Hands-AI/OpenHands?tab=readme-ov-file#-how-to-join-the-community)
-
-## 💪 Becoming a Contributor
-
-We welcome contributions from everyone! Whether you're a developer, a researcher, or simply enthusiastic about advancing
-the field of software engineering with AI, there are many ways to get involved:
-
-- **Code Contributions:** Help us develop new core functionality, improve our agents, improve the frontend and other
-interfaces, or anything else that would help make OpenHands better.
-- **Research and Evaluation:** Contribute to our understanding of LLMs in software engineering, participate in
-evaluating the models, or suggest improvements.
-- **Feedback and Testing:** Use the OpenHands toolset, report bugs, suggest features, or provide feedback on usability.
-
-For details, please check [CONTRIBUTING.md](./CONTRIBUTING.md).
-
-## Code of Conduct
-
-We have a [Code of Conduct](./CODE_OF_CONDUCT.md) that we expect all contributors to adhere to.
-Long story short, we are aiming for an open, welcoming, diverse, inclusive, and healthy community.
-All contributors are expected to contribute to building this sort of community.
-
-## 🛠️ Becoming a Maintainer
-
-For contributors who have made significant and sustained contributions to the project, there is a possibility of joining
-the maintainer team. The process for this is as follows:
-
-1. Any contributor who has made sustained and high-quality contributions to the codebase can be nominated by any
-maintainer. If you feel that you may qualify you can reach out to any of the maintainers that have reviewed your PRs and ask if you can be nominated.
-2. Once a maintainer nominates a new maintainer, there will be a discussion period among the maintainers for at least 3 days.
-3. If no concerns are raised the nomination will be accepted by acclamation, and if concerns are raised there will be a discussion and possible vote.
-
-Note that just making many PRs does not immediately imply that you will become a maintainer. We will be looking
-at sustained high-quality contributions over a period of time, as well as good teamwork and adherence to our [Code of Conduct](./CODE_OF_CONDUCT.md).

CONTRIBUTING.md

@@ -1,123 +0,0 @@
-# Contributing
-
-Thanks for your interest in contributing to OpenHands! We welcome and appreciate contributions.
-
-## Understanding OpenHands's CodeBase
-
-To understand the codebase, please refer to the README in each module:
-- [frontend](./frontend/README.md)
-- [evaluation](./evaluation/README.md)
-- [openhands](./openhands/README.md)
-   - [agenthub](./openhands/agenthub/README.md)
-   - [server](./openhands/server/README.md)
-
-## Setting up Your Development Environment
-
-We have a separate doc [Development.md](https://github.com/All-Hands-AI/OpenHands/blob/main/Development.md) that tells you how to set up a development workflow.
-
-## How Can I Contribute?
-
-There are many ways that you can contribute:
-
-1. **Download and use** OpenHands, and send [issues](https://github.com/All-Hands-AI/OpenHands/issues) when you encounter something that isn't working or a feature that you'd like to see.
-2. **Send feedback** after each session by [clicking the thumbs-up thumbs-down buttons](https://docs.all-hands.dev/modules/usage/feedback), so we can see where things are working and failing, and also build an open dataset for training code agents.
-3. **Improve the Codebase** by sending [PRs](#sending-pull-requests-to-openhands) (see details below). In particular, we have some [good first issues](https://github.com/All-Hands-AI/OpenHands/labels/good%20first%20issue) that may be ones to start on.
-
-## What Can I Build?
-Here are a few ways you can help improve the codebase.
-
-#### UI/UX
-We're always looking to improve the look and feel of the application. If you've got a small fix
-for something that's bugging you, feel free to open up a PR that changes the [`./frontend`](./frontend) directory.
-
-If you're looking to make a bigger change, add a new UI element, or significantly alter the style
-of the application, please open an issue first, or better, join the #frontend channel in our Slack
-to gather consensus from our design team first.
-
-#### Improving the agent
-Our main agent is the CodeAct agent. You can [see its prompts here](https://github.com/All-Hands-AI/OpenHands/tree/main/openhands/agenthub/codeact_agent).
-
-Changes to these prompts, and to the underlying behavior in Python, can have a huge impact on user experience.
-You can try modifying the prompts to see how they change the behavior of the agent as you use the app
-locally, but we will need to do an end-to-end evaluation of any changes here to ensure that the agent
-is getting better over time.
-
-We use the [SWE-bench](https://www.swebench.com/) benchmark to test our agent. You can join the #evaluation
-channel in Slack to learn more.
-
-#### Adding a new agent
-You may want to experiment with building new types of agents. You can add an agent to [`openhands/agenthub`](./openhands/agenthub)
-to help expand the capabilities of OpenHands.
-
-#### Adding a new runtime
-The agent needs a place to run code and commands. When you run OpenHands on your laptop, it uses a Docker container
-to do this by default. But there are other ways of creating a sandbox for the agent.
-
-If you work for a company that provides a cloud-based runtime, you could help us add support for that runtime
-by implementing the [interface specified here](https://github.com/All-Hands-AI/OpenHands/blob/main/openhands/runtime/base.py).
-
-#### Testing
-When you write code, it is also good to write tests. Please navigate to the [`./tests`](./tests) folder to see existing test suites.
-At the moment, we have two kinds of tests: [`unit`](./tests/unit) and [`integration`](./evaluation/integration_tests). Please refer to the README for each test suite. These tests also run on GitHub's continuous integration to ensure quality of the project.
-
-## Sending Pull Requests to OpenHands
-
-You'll need to fork our repository to send us a Pull Request. You can learn more
-about how to fork a GitHub repo and open a PR with your changes in [this article](https://medium.com/swlh/forks-and-pull-requests-how-to-contribute-to-github-repos-8843fac34ce8).
-
-### Pull Request title
-As described [here](https://github.com/commitizen/conventional-commit-types/blob/master/index.json), a valid PR title should begin with one of the following prefixes:
-
-- `feat`: A new feature
-- `fix`: A bug fix
-- `docs`: Documentation only changes
-- `style`: Changes that do not affect the meaning of the code (white space, formatting, missing semicolons, etc.)
-- `refactor`: A code change that neither fixes a bug nor adds a feature
-- `perf`: A code change that improves performance
-- `test`: Adding missing tests or correcting existing tests
-- `build`: Changes that affect the build system or external dependencies (example scopes: gulp, broccoli, npm)
-- `ci`: Changes to our CI configuration files and scripts (example scopes: Travis, Circle, BrowserStack, SauceLabs)
-- `chore`: Other changes that don't modify src or test files
-- `revert`: Reverts a previous commit
-
-For example, a PR title could be:
-- `refactor: modify package path`
-- `feat(frontend): xxxx`, where `(frontend)` means that this PR mainly focuses on the frontend component.
-
-You may also check out previous PRs in the [PR list](https://github.com/All-Hands-AI/OpenHands/pulls).
-
-### Pull Request description
-- If your PR is small (such as a typo fix), you can go brief.
-- If it contains a lot of changes, it's better to write more details.
-
-If your changes are user-facing (e.g. a new feature in the UI, a change in behavior, or a bugfix)
-please include a short message that we can add to our changelog.
-
-## How to Make Effective Contributions
-
-### Opening Issues
-
-If you notice any bugs or have any feature requests please open them via the [issues page](https://github.com/All-Hands-AI/OpenHands/issues). We will triage based on how critical the bug is or how potentially useful the improvement is, discuss, and implement the ones that the community has interest/effort for.
-
-Further, if you see an issue you like, please leave a "thumbs-up" or a comment, which will help us prioritize.
-
-### Making Pull Requests
-
-We're generally happy to consider all pull requests with the evaluation process varying based on the type of change:
-
-#### For Small Improvements
-
-Small improvements with few downsides are typically reviewed and approved quickly.
-One thing to check when making changes is to ensure that all continuous integration tests pass, which you can check before getting a review.
-
-#### For Core Agent Changes
-
-We need to be more careful with changes to the core agent, as it is imperative to maintain high quality. These PRs are evaluated based on three key metrics:
-
-1. **Accuracy**
-2. **Efficiency**
-3. **Code Complexity**
-
-If it improves accuracy, efficiency, or both with only a minimal change to code quality, that's great we're happy to merge it in!
-If there are bigger tradeoffs (e.g. helping efficiency a lot and hurting accuracy a little) we might want to put it behind a feature flag.
-Either way, please feel free to discuss on github issues or slack, and we will give guidance and preliminary feedback.

CREDITS.md

@@ -1,312 +0,0 @@
-# Credits
-
-## Contributors
-
-We would like to thank all the [contributors](https://github.com/All-Hands-AI/OpenHands/graphs/contributors) who have helped make OpenHands possible. We greatly appreciate your dedication and hard work.
-
-## Open Source Projects
-
-OpenHands includes and adapts the following open source projects. We are grateful for their contributions to the open source community:
-
-#### [SWE Agent](https://github.com/princeton-nlp/swe-agent)
-   - License: MIT License
-   - Description: Adapted for use in OpenHands's agent hub
-
-#### [Aider](https://github.com/paul-gauthier/aider)
-   - License: Apache License 2.0
-   - Description: AI pair programming tool. OpenHands has adapted and integrated its linter module for code-related tasks in [`agentskills utilities`](https://github.com/All-Hands-AI/OpenHands/tree/main/openhands/runtime/plugins/agent_skills/utils/aider)
-
-#### [BrowserGym](https://github.com/ServiceNow/BrowserGym)
-   - License: Apache License 2.0
-   - Description: Adapted in implementing the browsing agent
-
-
-### Reference Implementations for Evaluation Benchmarks
-OpenHands integrates code of the reference implementations for the following agent evaluation benchmarks:
-
-#### [HumanEval](https://github.com/openai/human-eval)
-   - License: MIT License
-
-#### [DSP](https://github.com/microsoft/DataScienceProblems)
-   - License: MIT License
-
-#### [HumanEvalPack](https://github.com/bigcode-project/bigcode-evaluation-harness)
-   - License: Apache License 2.0
-
-#### [AgentBench](https://github.com/THUDM/AgentBench)
-   - License: Apache License 2.0
-
-#### [SWE-Bench](https://github.com/princeton-nlp/SWE-bench)
-   - License: MIT License
-
-#### [BIRD](https://bird-bench.github.io/)
-   - License: MIT License
-   - Dataset: CC-BY-SA 4.0
-
-#### [Gorilla APIBench](https://github.com/ShishirPatil/gorilla)
-   - License: Apache License 2.0
-
-#### [GPQA](https://github.com/idavidrein/gpqa)
-   - License: MIT License
-
-#### [ProntoQA](https://github.com/asaparov/prontoqa)
-   - License: Apache License 2.0
-
-
-## Open Source licenses
-
-### MIT License
-
-Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
-
-The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
-
-THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
-
-### BSD 3-Clause License
-
-Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
-
-1. Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
-
-2. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
-
-3. Neither the name of the copyright holder nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.
-
-THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
-
-### Apache License 2.0
-
-
-                                 Apache License
-                           Version 2.0, January 2004
-                        http://www.apache.org/licenses/
-
-   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
-
-   1. Definitions.
-
-      "License" shall mean the terms and conditions for use, reproduction,
-      and distribution as defined by Sections 1 through 9 of this document.
-
-      "Licensor" shall mean the copyright owner or entity authorized by
-      the copyright owner that is granting the License.
-
-      "Legal Entity" shall mean the union of the acting entity and all
-      other entities that control, are controlled by, or are under common
-      control with that entity. For the purposes of this definition,
-      "control" means (i) the power, direct or indirect, to cause the
-      direction or management of such entity, whether by contract or
-      otherwise, or (ii) ownership of fifty percent (50%) or more of the
-      outstanding shares, or (iii) beneficial ownership of such entity.
-
-      "You" (or "Your") shall mean an individual or Legal Entity
-      exercising permissions granted by this License.
-
-      "Source" form shall mean the preferred form for making modifications,
-      including but not limited to software source code, documentation
-      source, and configuration files.
-
-      "Object" form shall mean any form resulting from mechanical
-      transformation or translation of a Source form, including but
-      not limited to compiled object code, generated documentation,
-      and conversions to other media types.
-
-      "Work" shall mean the work of authorship, whether in Source or
-      Object form, made available under the License, as indicated by a
-      copyright notice that is included in or attached to the work
-      (an example is provided in the Appendix below).
-
-      "Derivative Works" shall mean any work, whether in Source or Object
-      form, that is based on (or derived from) the Work and for which the
-      editorial revisions, annotations, elaborations, or other modifications
-      represent, as a whole, an original work of authorship. For the purposes
-      of this License, Derivative Works shall not include works that remain
-      separable from, or merely link (or bind by name) to the interfaces of,
-      the Work and Derivative Works thereof.
-
-      "Contribution" shall mean any work of authorship, including
-      the original version of the Work and any modifications or additions
-      to that Work or Derivative Works thereof, that is intentionally
-      submitted to Licensor for inclusion in the Work by the copyright owner
-      or by an individual or Legal Entity authorized to submit on behalf of
-      the copyright owner. For the purposes of this definition, "submitted"
-      means any form of electronic, verbal, or written communication sent
-      to the Licensor or its representatives, including but not limited to
-      communication on electronic mailing lists, source code control systems,
-      and issue tracking systems that are managed by, or on behalf of, the
-      Licensor for the purpose of discussing and improving the Work, but
-      excluding communication that is conspicuously marked or otherwise
-      designated in writing by the copyright owner as "Not a Contribution."
-
-      "Contributor" shall mean Licensor and any individual or Legal Entity
-      on behalf of whom a Contribution has been received by Licensor and
-      subsequently incorporated within the Work.
-
-   2. Grant of Copyright License. Subject to the terms and conditions of
-      this License, each Contributor hereby grants to You a perpetual,
-      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
-      copyright license to reproduce, prepare Derivative Works of,
-      publicly display, publicly perform, sublicense, and distribute the
-      Work and such Derivative Works in Source or Object form.
-
-   3. Grant of Patent License. Subject to the terms and conditions of
-      this License, each Contributor hereby grants to You a perpetual,
-      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
-      (except as stated in this section) patent license to make, have made,
-      use, offer to sell, sell, import, and otherwise transfer the Work,
-      where such license applies only to those patent claims licensable
-      by such Contributor that are necessarily infringed by their
-      Contribution(s) alone or by combination of their Contribution(s)
-      with the Work to which such Contribution(s) was submitted. If You
-      institute patent litigation against any entity (including a
-      cross-claim or counterclaim in a lawsuit) alleging that the Work
-      or a Contribution incorporated within the Work constitutes direct
-      or contributory patent infringement, then any patent licenses
-      granted to You under this License for that Work shall terminate
-      as of the date such litigation is filed.
-
-   4. Redistribution. You may reproduce and distribute copies of the
-      Work or Derivative Works thereof in any medium, with or without
-      modifications, and in Source or Object form, provided that You
-      meet the following conditions:
-
-      (a) You must give any other recipients of the Work or
-          Derivative Works a copy of this License; and
-
-      (b) You must cause any modified files to carry prominent notices
-          stating that You changed the files; and
-
-      (c) You must retain, in the Source form of any Derivative Works
-          that You distribute, all copyright, patent, trademark, and
-          attribution notices from the Source form of the Work,
-          excluding those notices that do not pertain to any part of
-          the Derivative Works; and
-
-      (d) If the Work includes a "NOTICE" text file as part of its
-          distribution, then any Derivative Works that You distribute must
-          include a readable copy of the attribution notices contained
-          within such NOTICE file, excluding those notices that do not
-          pertain to any part of the Derivative Works, in at least one
-          of the following places: within a NOTICE text file distributed
-          as part of the Derivative Works; within the Source form or
-          documentation, if provided along with the Derivative Works; or,
-          within a display generated by the Derivative Works, if and
-          wherever such third-party notices normally appear. The contents
-          of the NOTICE file are for informational purposes only and
-          do not modify the License. You may add Your own attribution
-          notices within Derivative Works that You distribute, alongside
-          or as an addendum to the NOTICE text from the Work, provided
-          that such additional attribution notices cannot be construed
-          as modifying the License.
-
-      You may add Your own copyright statement to Your modifications and
-      may provide additional or different license terms and conditions
-      for use, reproduction, or distribution of Your modifications, or
-      for any such Derivative Works as a whole, provided Your use,
-      reproduction, and distribution of the Work otherwise complies with
-      the conditions stated in this License.
-
-   5. Submission of Contributions. Unless You explicitly state otherwise,
-      any Contribution intentionally submitted for inclusion in the Work
-      by You to the Licensor shall be under the terms and conditions of
-      this License, without any additional terms or conditions.
-      Notwithstanding the above, nothing herein shall supersede or modify
-      the terms of any separate license agreement you may have executed
-      with Licensor regarding such Contributions.
-
-   6. Trademarks. This License does not grant permission to use the trade
-      names, trademarks, service marks, or product names of the Licensor,
-      except as required for reasonable and customary use in describing the
-      origin of the Work and reproducing the content of the NOTICE file.
-
-   7. Disclaimer of Warranty. Unless required by applicable law or
-      agreed to in writing, Licensor provides the Work (and each
-      Contributor provides its Contributions) on an "AS IS" BASIS,
-      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
-      implied, including, without limitation, any warranties or conditions
-      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
-      PARTICULAR PURPOSE. You are solely responsible for determining the
-      appropriateness of using or redistributing the Work and assume any
-      risks associated with Your exercise of permissions under this License.
-
-   8. Limitation of Liability. In no event and under no legal theory,
-      whether in tort (including negligence), contract, or otherwise,
-      unless required by applicable law (such as deliberate and grossly
-      negligent acts) or agreed to in writing, shall any Contributor be
-      liable to You for damages, including any direct, indirect, special,
-      incidental, or consequential damages of any character arising as a
-      result of this License or out of the use or inability to use the
-      Work (including but not limited to damages for loss of goodwill,
-      work stoppage, computer failure or malfunction, or any and all
-      other commercial damages or losses), even if such Contributor
-      has been advised of the possibility of such damages.
-
-   9. Accepting Warranty or Additional Liability. While redistributing
-      the Work or Derivative Works thereof, You may choose to offer,
-      and charge a fee for, acceptance of support, warranty, indemnity,
-      or other liability obligations and/or rights consistent with this
-      License. However, in accepting such obligations, You may act only
-      on Your own behalf and on Your sole responsibility, not on behalf
-      of any other Contributor, and only if You agree to indemnify,
-      defend, and hold each Contributor harmless for any liability
-      incurred by, or claims asserted against, such Contributor by reason
-      of your accepting any such warranty or additional liability.
-
-   END OF TERMS AND CONDITIONS
-
-   APPENDIX: How to apply the Apache License to your work.
-
-      To apply the Apache License to your work, attach the following
-      boilerplate notice, with the fields enclosed by brackets "[]"
-      replaced with your own identifying information. (Don't include
-      the brackets!)  The text should be enclosed in the appropriate
-      comment syntax for the file format. We also recommend that a
-      file or class name and description of purpose be included on the
-      same "printed page" as the copyright notice for easier
-      identification within third-party archives.
-
-   Copyright [yyyy] [name of copyright owner]
-
-
-
-### Non-Open Source Reference Implementations:
-
-#### [MultiPL-E](https://github.com/nuprl/MultiPL-E)
-   - License: BSD 3-Clause License with Machine Learning Restriction
-
-BSD 3-Clause License with Machine Learning Restriction
-
-Copyright (c) 2022, Northeastern University, Oberlin College, Roblox Inc,
-Stevens Institute of Technology, University of Massachusetts Amherst, and
-Wellesley College.
-
-All rights reserved.
-
-Redistribution and use in source and binary forms, with or without
-modification, are permitted provided that the following conditions are met:
-
-1. Redistributions of source code must retain the above copyright notice, this
-   list of conditions and the following disclaimer.
-
-2. Redistributions in binary form must reproduce the above copyright notice,
-   this list of conditions and the following disclaimer in the documentation
-   and/or other materials provided with the distribution.
-
-3. Neither the name of the copyright holder nor the names of its
-   contributors may be used to endorse or promote products derived from
-   this software without specific prior written permission.
-
-4.  The contents of this repository may not be used as training data for any
-    machine learning model, including but not limited to neural networks.
-
-THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
-AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
-IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
-DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
-FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
-DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
-SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
-CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
-OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
-OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

Development.md

@@ -1,128 +0,0 @@
-# Development Guide
-This guide is for people working on OpenHands and editing the source code.
-If you wish to contribute your changes, check out the [CONTRIBUTING.md](https://github.com/All-Hands-AI/OpenHands/blob/main/CONTRIBUTING.md) on how to clone and setup the project initially before moving on.
-Otherwise, you can clone the OpenHands project directly.
-
-## Start the Server for Development
-### 1. Requirements
-* Linux, Mac OS, or [WSL on Windows](https://learn.microsoft.com/en-us/windows/wsl/install)  [Ubuntu >= 22.04]
-* [Docker](https://docs.docker.com/engine/install/) (For those on MacOS, make sure to allow the default Docker socket to be used from advanced settings!)
-* [Python](https://www.python.org/downloads/) = 3.12
-* [NodeJS](https://nodejs.org/en/download/package-manager) >= 20.x
-* [Poetry](https://python-poetry.org/docs/#installing-with-the-official-installer) >= 1.8
-* OS-specific dependencies:
-  - Ubuntu: build-essential => `sudo apt-get install build-essential`
-  - WSL: netcat => `sudo apt-get install netcat`
-
-Make sure you have all these dependencies installed before moving on to `make build`.
-
-#### Develop without sudo access
-If you want to develop without system admin/sudo access to upgrade/install `Python` and/or `NodeJs`, you can use `conda` or `mamba` to manage the packages for you:
-
-```bash
-# Download and install Mamba (a faster version of conda)
-curl -L -O "https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-$(uname)-$(uname -m).sh"
-bash Miniforge3-$(uname)-$(uname -m).sh
-
-# Install Python 3.12, nodejs, and poetry
-mamba install python=3.12
-mamba install conda-forge::nodejs
-mamba install conda-forge::poetry
-```
-
-### 2. Build and Setup The Environment
-Begin by building the project which includes setting up the environment and installing dependencies. This step ensures that OpenHands is ready to run on your system:
-
-```bash
-make build
-```
-
-### 3. Configuring the Language Model
-OpenHands supports a diverse array of Language Models (LMs) through the powerful [litellm](https://docs.litellm.ai) library.
-By default, we've chosen Claude Sonnet 3.5 as our go-to model, but the world is your oyster! You can unleash the
-potential of any other LM that piques your interest.
-
-To configure the LM of your choice, run:
-
-   ```bash
-   make setup-config
-   ```
-
-   This command will prompt you to enter the LLM API key, model name, and other variables ensuring that OpenHands is tailored to your specific needs. Note that the model name will apply only when you run headless. If you use the UI, please set the model in the UI.
-
-   Note: If you have previously run OpenHands using the docker command, you may have already set some environmental variables in your terminal. The final configurations are set from highest to lowest priority:
-   Environment variables > config.toml variables > default variables
-
-**Note on Alternative Models:**
-See [our documentation](https://docs.all-hands.dev/modules/usage/llms) for recommended models.
-
-### 4. Running the application
-#### Option A: Run the Full Application
-Once the setup is complete, this command starts both the backend and frontend servers, allowing you to interact with OpenHands:
-```bash
-make run
-```
-
-#### Option B: Individual Server Startup
-- **Start the Backend Server:** If you prefer, you can start the backend server independently to focus on backend-related tasks or configurations.
-    ```bash
-    make start-backend
-    ```
-
-- **Start the Frontend Server:** Similarly, you can start the frontend server on its own to work on frontend-related components or interface enhancements.
-    ```bash
-    make start-frontend
-    ```
-
-### 6. LLM Debugging
-If you encounter any issues with the Language Model (LM) or you're simply curious, export DEBUG=1 in the environment and restart the backend.
-OpenHands will log the prompts and responses in the logs/llm/CURRENT_DATE directory, allowing you to identify the causes.
-
-### 7. Help
-Need help or info on available targets and commands? Use the help command for all the guidance you need with OpenHands.
-```bash
-make help
- ```
-
-### 8. Testing
-To run tests, refer to the following:
-#### Unit tests
-
-```bash
-poetry run pytest ./tests/unit/test_*.py
-```
-
-### 9. Add or update dependency
-1. Add your dependency in `pyproject.toml` or use `poetry add xxx`.
-2. Update the poetry.lock file via `poetry lock --no-update`.
-
-### 9. Use existing Docker image
-To reduce build time (e.g., if no changes were made to the client-runtime component), you can use an existing Docker container image by
-setting the SANDBOX_RUNTIME_CONTAINER_IMAGE environment variable to the desired Docker image.
-
-Example: `export SANDBOX_RUNTIME_CONTAINER_IMAGE=ghcr.io/all-hands-ai/runtime:0.27-nikolaik`
-
-## Develop inside Docker container
-
-TL;DR
-
-```bash
-make docker-dev
-```
-
-See more details [here](./containers/dev/README.md).
-
-If you are just interested in running `OpenHands` without installing all the required tools on your host.
-
-```bash
-make docker-run
-```
-
-If you do not have `make` on your host, run:
-
-```bash
-cd ./containers/dev
-./dev.sh
-```
-
-You do need [Docker](https://docs.docker.com/engine/install/) installed on your host though.

ISSUE_TRIAGE.md

@@ -1,32 +0,0 @@
-# Issue Triage
-These are the procedures and guidelines on how issues are triaged in this repo by the maintainers.
-
-## General
-* All issues must be tagged with **enhancement**, **bug** or **troubleshooting/help**.
-* Issues may be tagged with what it relates to (**agent quality**, **frontend**, **resolver**, etc.).
-
-## Severity
-* **Low**: Minor issues or affecting single user.
-* **Medium**: Affecting multiple users.
-* **High**: High visibility issues or affecting many users.
-* **Critical**: Affecting all users or potential security issues.
-
-## Effort
-* Issues may be estimated with effort required (**small effort**, **medium effort**, **large effort**).
-
-## Difficulty
-* Issues with low implementation difficulty may be tagged with **good first issue**.
-
-## Not Enough Information
-* User is asked to provide more information (logs, how to reproduce, etc.) when the issue is not clear.
-* If an issue is unclear and the author does not provide more information or respond to a request,
-the issue may be closed as **not planned** (Usually after a week).
-
-## Multiple Requests/Fixes in One Issue
-* These issues will be narrowed down to one request/fix so the issue is more easily tracked and fixed.
-* Issues may be broken down into multiple issues if required.
-
-## Stale and Auto Closures
-* In order to keep a maintainable backlog, issues that have no activity within 30 days are automatically marked as **Stale**.
-* If issues marked as **Stale** continue to have no activity for 7 more days, they will automatically be closed as not planned.
-* Issues may be reopened by maintainers if deemed important.

LICENSE

@@ -1,25 +0,0 @@
-The MIT License (MIT)
-=====================
-
-Copyright © 2023
-
-Permission is hereby granted, free of charge, to any person
-obtaining a copy of this software and associated documentation
-files (the “Software”), to deal in the Software without
-restriction, including without limitation the rights to use,
-copy, modify, merge, publish, distribute, sublicense, and/or sell
-copies of the Software, and to permit persons to whom the
-Software is furnished to do so, subject to the following
-conditions:
-
-The above copyright notice and this permission notice shall be
-included in all copies or substantial portions of the Software.
-
-THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND,
-EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES
-OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
-NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
-HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
-WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
-FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
-OTHER DEALINGS IN THE SOFTWARE.

MANIFEST.in

@@ -1,5 +0,0 @@
-# Exclude all Python bytecode files
-global-exclude *.pyc
-
-# Exclude Python cache directories
-global-exclude __pycache__

Makefile

@@ -1,332 +0,0 @@
-SHELL=/usr/bin/env bash
-# Makefile for OpenHands project
-
-# Variables
-BACKEND_HOST ?= "127.0.0.1"
-BACKEND_PORT = 3000
-BACKEND_HOST_PORT = "$(BACKEND_HOST):$(BACKEND_PORT)"
-FRONTEND_PORT = 3001
-DEFAULT_WORKSPACE_DIR = "./workspace"
-DEFAULT_MODEL = "gpt-4o"
-CONFIG_FILE = config.toml
-PRE_COMMIT_CONFIG_PATH = "./dev_config/python/.pre-commit-config.yaml"
-PYTHON_VERSION = 3.12
-
-# ANSI color codes
-GREEN=$(shell tput -Txterm setaf 2)
-YELLOW=$(shell tput -Txterm setaf 3)
-RED=$(shell tput -Txterm setaf 1)
-BLUE=$(shell tput -Txterm setaf 6)
-RESET=$(shell tput -Txterm sgr0)
-
-# Build
-build:
-	@echo "$(GREEN)Building project...$(RESET)"
-	@$(MAKE) -s check-dependencies
-	@$(MAKE) -s install-python-dependencies
-	@$(MAKE) -s install-frontend-dependencies
-	@$(MAKE) -s install-pre-commit-hooks
-	@$(MAKE) -s build-frontend
-	@echo "$(GREEN)Build completed successfully.$(RESET)"
-
-check-dependencies:
-	@echo "$(YELLOW)Checking dependencies...$(RESET)"
-	@$(MAKE) -s check-system
-	@$(MAKE) -s check-python
-	@$(MAKE) -s check-npm
-	@$(MAKE) -s check-nodejs
-ifeq ($(INSTALL_DOCKER),)
-	@$(MAKE) -s check-docker
-endif
-	@$(MAKE) -s check-poetry
-	@echo "$(GREEN)Dependencies checked successfully.$(RESET)"
-
-check-system:
-	@echo "$(YELLOW)Checking system...$(RESET)"
-	@if [ "$(shell uname)" = "Darwin" ]; then \
-		echo "$(BLUE)macOS detected.$(RESET)"; \
-	elif [ "$(shell uname)" = "Linux" ]; then \
-		if [ -f "/etc/manjaro-release" ]; then \
-			echo "$(BLUE)Manjaro Linux detected.$(RESET)"; \
-		else \
-			echo "$(BLUE)Linux detected.$(RESET)"; \
-		fi; \
-	elif [ "$$(uname -r | grep -i microsoft)" ]; then \
-		echo "$(BLUE)Windows Subsystem for Linux detected.$(RESET)"; \
-	else \
-		echo "$(RED)Unsupported system detected. Please use macOS, Linux, or Windows Subsystem for Linux (WSL).$(RESET)"; \
-		exit 1; \
-	fi
-
-check-python:
-	@echo "$(YELLOW)Checking Python installation...$(RESET)"
-	@if command -v python$(PYTHON_VERSION) > /dev/null; then \
-		echo "$(BLUE)$(shell python$(PYTHON_VERSION) --version) is already installed.$(RESET)"; \
-	else \
-		echo "$(RED)Python $(PYTHON_VERSION) is not installed. Please install Python $(PYTHON_VERSION) to continue.$(RESET)"; \
-		exit 1; \
-	fi
-
-check-npm:
-	@echo "$(YELLOW)Checking npm installation...$(RESET)"
-	@if command -v npm > /dev/null; then \
-		echo "$(BLUE)npm $(shell npm --version) is already installed.$(RESET)"; \
-	else \
-		echo "$(RED)npm is not installed. Please install Node.js to continue.$(RESET)"; \
-		exit 1; \
-	fi
-
-check-nodejs:
-	@echo "$(YELLOW)Checking Node.js installation...$(RESET)"
-	@if command -v node > /dev/null; then \
-		NODE_VERSION=$(shell node --version | sed -E 's/v//g'); \
-		IFS='.' read -r -a NODE_VERSION_ARRAY <<< "$$NODE_VERSION"; \
-		if [ "$${NODE_VERSION_ARRAY[0]}" -ge 22 ]; then \
-			echo "$(BLUE)Node.js $$NODE_VERSION is already installed.$(RESET)"; \
-		else \
-			echo "$(RED)Node.js 22.x or later is required. Please install Node.js 22.x or later to continue.$(RESET)"; \
-			exit 1; \
-		fi; \
-	else \
-		echo "$(RED)Node.js is not installed. Please install Node.js to continue.$(RESET)"; \
-		exit 1; \
-	fi
-
-check-docker:
-	@echo "$(YELLOW)Checking Docker installation...$(RESET)"
-	@if command -v docker > /dev/null; then \
-		echo "$(BLUE)$(shell docker --version) is already installed.$(RESET)"; \
-	else \
-		echo "$(RED)Docker is not installed. Please install Docker to continue.$(RESET)"; \
-		exit 1; \
-	fi
-
-check-poetry:
-	@echo "$(YELLOW)Checking Poetry installation...$(RESET)"
-	@if command -v poetry > /dev/null; then \
-		POETRY_VERSION=$(shell poetry --version 2>&1 | sed -E 's/Poetry \(version ([0-9]+\.[0-9]+\.[0-9]+)\)/\1/'); \
-		IFS='.' read -r -a POETRY_VERSION_ARRAY <<< "$$POETRY_VERSION"; \
-		if [ $${POETRY_VERSION_ARRAY[0]} -gt 1 ] || ([ $${POETRY_VERSION_ARRAY[0]} -eq 1 ] && [ $${POETRY_VERSION_ARRAY[1]} -ge 8 ]); then \
-			echo "$(BLUE)$(shell poetry --version) is already installed.$(RESET)"; \
-		else \
-			echo "$(RED)Poetry 1.8 or later is required. You can install poetry by running the following command, then adding Poetry to your PATH:"; \
-			echo "$(RED) curl -sSL https://install.python-poetry.org | python$(PYTHON_VERSION) -$(RESET)"; \
-			echo "$(RED)More detail here: https://python-poetry.org/docs/#installing-with-the-official-installer$(RESET)"; \
-			exit 1; \
-		fi; \
-	else \
-		echo "$(RED)Poetry is not installed. You can install poetry by running the following command, then adding Poetry to your PATH:"; \
-		echo "$(RED) curl -sSL https://install.python-poetry.org | python$(PYTHON_VERSION) -$(RESET)"; \
-		echo "$(RED)More detail here: https://python-poetry.org/docs/#installing-with-the-official-installer$(RESET)"; \
-		exit 1; \
-	fi
-
-install-python-dependencies:
-	@echo "$(GREEN)Installing Python dependencies...$(RESET)"
-	@if [ -z "${TZ}" ]; then \
-		echo "Defaulting TZ (timezone) to UTC"; \
-		export TZ="UTC"; \
-	fi
-	poetry env use python$(PYTHON_VERSION)
-	@if [ "$(shell uname)" = "Darwin" ]; then \
-		echo "$(BLUE)Installing chroma-hnswlib...$(RESET)"; \
-		export HNSWLIB_NO_NATIVE=1; \
-		poetry run pip install chroma-hnswlib; \
-	fi
-	@poetry install --without llama-index
-	@if [ -f "/etc/manjaro-release" ]; then \
-		echo "$(BLUE)Detected Manjaro Linux. Installing Playwright dependencies...$(RESET)"; \
-		poetry run pip install playwright; \
-		poetry run playwright install chromium; \
-	else \
-		if [ ! -f cache/playwright_chromium_is_installed.txt ]; then \
-			echo "Running playwright install --with-deps chromium..."; \
-			poetry run playwright install --with-deps chromium; \
-			mkdir -p cache; \
-			touch cache/playwright_chromium_is_installed.txt; \
-		else \
-			echo "Setup already done. Skipping playwright installation."; \
-		fi \
-	fi
-	@echo "$(GREEN)Python dependencies installed successfully.$(RESET)"
-
-install-frontend-dependencies:
-	@echo "$(YELLOW)Setting up frontend environment...$(RESET)"
-	@echo "$(YELLOW)Detect Node.js version...$(RESET)"
-	@cd frontend && node ./scripts/detect-node-version.js
-	echo "$(BLUE)Installing frontend dependencies with npm...$(RESET)"
-	@cd frontend && npm install
-	@echo "$(GREEN)Frontend dependencies installed successfully.$(RESET)"
-
-install-pre-commit-hooks:
-	@echo "$(YELLOW)Installing pre-commit hooks...$(RESET)"
-	@git config --unset-all core.hooksPath || true
-	@poetry run pre-commit install --config $(PRE_COMMIT_CONFIG_PATH)
-	@echo "$(GREEN)Pre-commit hooks installed successfully.$(RESET)"
-
-lint-backend:
-	@echo "$(YELLOW)Running linters...$(RESET)"
-	@poetry run pre-commit run --files openhands/**/* agenthub/**/* evaluation/**/* --show-diff-on-failure --config $(PRE_COMMIT_CONFIG_PATH)
-
-lint-frontend:
-	@echo "$(YELLOW)Running linters for frontend...$(RESET)"
-	@cd frontend && npm run lint
-
-lint:
-	@$(MAKE) -s lint-frontend
-	@$(MAKE) -s lint-backend
-
-test-frontend:
-	@echo "$(YELLOW)Running tests for frontend...$(RESET)"
-	@cd frontend && npm run test
-
-test:
-	@$(MAKE) -s test-frontend
-
-build-frontend:
-	@echo "$(YELLOW)Building frontend...$(RESET)"
-	@cd frontend && npm run build
-
-# Start backend
-start-backend:
-	@echo "$(YELLOW)Starting backend...$(RESET)"
-	@poetry run uvicorn openhands.server.listen:app --host $(BACKEND_HOST) --port $(BACKEND_PORT) --reload --reload-exclude "./workspace"
-
-# Start frontend
-start-frontend:
-	@echo "$(YELLOW)Starting frontend...$(RESET)"
-	@cd frontend && VITE_BACKEND_HOST=$(BACKEND_HOST_PORT) VITE_FRONTEND_PORT=$(FRONTEND_PORT) npm run dev -- --port $(FRONTEND_PORT) --host $(BACKEND_HOST)
-
-# Common setup for running the app (non-callable)
-_run_setup:
-	@if [ "$(OS)" = "Windows_NT" ]; then \
-		echo "$(RED) Windows is not supported, use WSL instead!$(RESET)"; \
-		exit 1; \
-	fi
-	@mkdir -p logs
-	@echo "$(YELLOW)Starting backend server...$(RESET)"
-	@poetry run uvicorn openhands.server.listen:app --host $(BACKEND_HOST) --port $(BACKEND_PORT) &
-	@echo "$(YELLOW)Waiting for the backend to start...$(RESET)"
-	@until nc -z localhost $(BACKEND_PORT); do sleep 0.1; done
-	@echo "$(GREEN)Backend started successfully.$(RESET)"
-
-# Run the app (standard mode)
-run:
-	@echo "$(YELLOW)Running the app...$(RESET)"
-	@$(MAKE) -s _run_setup
-	@$(MAKE) -s start-frontend
-	@echo "$(GREEN)Application started successfully.$(RESET)"
-
-# Run the app (in docker)
-docker-run: WORKSPACE_BASE ?= $(PWD)/workspace
-docker-run:
-	@if [ -f /.dockerenv ]; then \
-		echo "Running inside a Docker container. Exiting..."; \
-		exit 0; \
-	else \
-		echo "$(YELLOW)Running the app in Docker $(OPTIONS)...$(RESET)"; \
-		export WORKSPACE_BASE=${WORKSPACE_BASE}; \
-		export SANDBOX_USER_ID=$(shell id -u); \
-		export DATE=$(shell date +%Y%m%d%H%M%S); \
-		docker compose up $(OPTIONS); \
-	fi
-
-# Run the app (WSL mode)
-run-wsl:
-	@echo "$(YELLOW)Running the app in WSL mode...$(RESET)"
-	@$(MAKE) -s _run_setup
-	@cd frontend && echo "$(BLUE)Starting frontend with npm (WSL mode)...$(RESET)" && npm run dev_wsl -- --port $(FRONTEND_PORT)
-	@echo "$(GREEN)Application started successfully in WSL mode.$(RESET)"
-
-# Setup config.toml
-setup-config:
-	@echo "$(YELLOW)Setting up config.toml...$(RESET)"
-	@$(MAKE) setup-config-prompts
-	@mv $(CONFIG_FILE).tmp $(CONFIG_FILE)
-	@echo "$(GREEN)Config.toml setup completed.$(RESET)"
-
-setup-config-prompts:
-	@echo "[core]" > $(CONFIG_FILE).tmp
-
-	@read -p "Enter your workspace directory (as absolute path) [default: $(DEFAULT_WORKSPACE_DIR)]: " workspace_dir; \
-	 workspace_dir=$${workspace_dir:-$(DEFAULT_WORKSPACE_DIR)}; \
-	 echo "workspace_base=\"$$workspace_dir\"" >> $(CONFIG_FILE).tmp
-
-	@echo "" >> $(CONFIG_FILE).tmp
-
-	@echo "[llm]" >> $(CONFIG_FILE).tmp
-	@read -p "Enter your LLM model name, used for running without UI. Set the model in the UI after you start the app. (see https://docs.litellm.ai/docs/providers for full list) [default: $(DEFAULT_MODEL)]: " llm_model; \
-	 llm_model=$${llm_model:-$(DEFAULT_MODEL)}; \
-	 echo "model=\"$$llm_model\"" >> $(CONFIG_FILE).tmp
-
-	@read -p "Enter your LLM api key: " llm_api_key; \
-	 echo "api_key=\"$$llm_api_key\"" >> $(CONFIG_FILE).tmp
-
-	@read -p "Enter your LLM base URL [mostly used for local LLMs, leave blank if not needed - example: http://localhost:5001/v1/]: " llm_base_url; \
-	 if [[ ! -z "$$llm_base_url" ]]; then echo "base_url=\"$$llm_base_url\"" >> $(CONFIG_FILE).tmp; fi
-
-	@echo "Enter your LLM Embedding Model"; \
-		echo "Choices are:"; \
-		echo "  - openai"; \
-		echo "  - azureopenai"; \
-		echo "  - Embeddings available only with OllamaEmbedding:"; \
-		echo "    - llama2"; \
-		echo "    - mxbai-embed-large"; \
-		echo "    - nomic-embed-text"; \
-		echo "    - all-minilm"; \
-		echo "    - stable-code"; \
-		echo "    - bge-m3"; \
-		echo "    - bge-large"; \
-		echo "    - paraphrase-multilingual"; \
-		echo "    - snowflake-arctic-embed"; \
-		echo "  - Leave blank to default to 'BAAI/bge-small-en-v1.5' via huggingface"; \
-		read -p "> " llm_embedding_model; \
-		echo "embedding_model=\"$$llm_embedding_model\"" >> $(CONFIG_FILE).tmp; \
-		if [ "$$llm_embedding_model" = "llama2" ] || [ "$$llm_embedding_model" = "mxbai-embed-large" ] || [ "$$llm_embedding_model" = "nomic-embed-text" ] || [ "$$llm_embedding_model" = "all-minilm" ] || [ "$$llm_embedding_model" = "stable-code" ]; then \
-			read -p "Enter the local model URL for the embedding model (will set llm.embedding_base_url): " llm_embedding_base_url; \
-				echo "embedding_base_url=\"$$llm_embedding_base_url\"" >> $(CONFIG_FILE).tmp; \
-		elif [ "$$llm_embedding_model" = "azureopenai" ]; then \
-			read -p "Enter the Azure endpoint URL (will overwrite llm.base_url): " llm_base_url; \
-				echo "base_url=\"$$llm_base_url\"" >> $(CONFIG_FILE).tmp; \
-			read -p "Enter the Azure LLM Embedding Deployment Name: " llm_embedding_deployment_name; \
-				echo "embedding_deployment_name=\"$$llm_embedding_deployment_name\"" >> $(CONFIG_FILE).tmp; \
-			read -p "Enter the Azure API Version: " llm_api_version; \
-				echo "api_version=\"$$llm_api_version\"" >> $(CONFIG_FILE).tmp; \
-		fi
-
-
-# Develop in container
-docker-dev:
-	@if [ -f /.dockerenv ]; then \
-		echo "Running inside a Docker container. Exiting..."; \
-		exit 0; \
-	else \
-		echo "$(YELLOW)Build and run in Docker $(OPTIONS)...$(RESET)"; \
-		./containers/dev/dev.sh $(OPTIONS); \
-	fi
-
-# Clean up all caches
-clean:
-	@echo "$(YELLOW)Cleaning up caches...$(RESET)"
-	@rm -rf openhands/.cache
-	@echo "$(GREEN)Caches cleaned up successfully.$(RESET)"
-
-# Help
-help:
-	@echo "$(BLUE)Usage: make [target]$(RESET)"
-	@echo "Targets:"
-	@echo "  $(GREEN)build$(RESET)               - Build project, including environment setup and dependencies."
-	@echo "  $(GREEN)lint$(RESET)                - Run linters on the project."
-	@echo "  $(GREEN)setup-config$(RESET)        - Setup the configuration for OpenHands by providing LLM API key,"
-	@echo "                        LLM Model name, and workspace directory."
-	@echo "  $(GREEN)start-backend$(RESET)       - Start the backend server for the OpenHands project."
-	@echo "  $(GREEN)start-frontend$(RESET)      - Start the frontend server for the OpenHands project."
-	@echo "  $(GREEN)run$(RESET)                 - Run the OpenHands application, starting both backend and frontend servers."
-	@echo "                        Backend Log file will be stored in the 'logs' directory."
-	@echo "  $(GREEN)docker-dev$(RESET)          - Build and run the OpenHands application in Docker."
-	@echo "  $(GREEN)docker-run$(RESET)          - Run the OpenHands application, starting both backend and frontend servers in Docker."
-	@echo "  $(GREEN)help$(RESET)                - Display this help message, providing information on available targets."
-
-# Phony targets
-.PHONY: build check-dependencies check-python check-npm check-docker check-poetry install-python-dependencies install-frontend-dependencies install-pre-commit-hooks lint start-backend start-frontend run run-wsl setup-config setup-config-prompts help
-.PHONY: docker-dev docker-run

README.md

@@ -1,137 +0,0 @@
-<a name="readme-top"></a>
-
-<div align="center">
-  <img src="./docs/static/img/logo.png" alt="Logo" width="200">
-  <h1 align="center">OpenHands: Code Less, Make More</h1>
-</div>
-
-
-<div align="center">
-  <a href="https://github.com/All-Hands-AI/OpenHands/graphs/contributors"><img src="https://img.shields.io/github/contributors/All-Hands-AI/OpenHands?style=for-the-badge&color=blue" alt="Contributors"></a>
-  <a href="https://github.com/All-Hands-AI/OpenHands/stargazers"><img src="https://img.shields.io/github/stars/All-Hands-AI/OpenHands?style=for-the-badge&color=blue" alt="Stargazers"></a>
-  <a href="https://codecov.io/github/All-Hands-AI/OpenHands?branch=main"><img alt="CodeCov" src="https://img.shields.io/codecov/c/github/All-Hands-AI/OpenHands?style=for-the-badge&color=blue"></a>
-  <a href="https://github.com/All-Hands-AI/OpenHands/blob/main/LICENSE"><img src="https://img.shields.io/github/license/All-Hands-AI/OpenHands?style=for-the-badge&color=blue" alt="MIT License"></a>
-  <br/>
-  <a href="https://join.slack.com/t/openhands-ai/shared_invite/zt-2ypg5jweb-d~6hObZDbXi_HEL8PDrbHg"><img src="https://img.shields.io/badge/Slack-Join%20Us-red?logo=slack&logoColor=white&style=for-the-badge" alt="Join our Slack community"></a>
-  <a href="https://discord.gg/ESHStjSjD4"><img src="https://img.shields.io/badge/Discord-Join%20Us-purple?logo=discord&logoColor=white&style=for-the-badge" alt="Join our Discord community"></a>
-  <a href="https://github.com/All-Hands-AI/OpenHands/blob/main/CREDITS.md"><img src="https://img.shields.io/badge/Project-Credits-blue?style=for-the-badge&color=FFE165&logo=github&logoColor=white" alt="Credits"></a>
-  <br/>
-  <a href="https://docs.all-hands.dev/modules/usage/getting-started"><img src="https://img.shields.io/badge/Documentation-000?logo=googledocs&logoColor=FFE165&style=for-the-badge" alt="Check out the documentation"></a>
-  <a href="https://arxiv.org/abs/2407.16741"><img src="https://img.shields.io/badge/Paper%20on%20Arxiv-000?logoColor=FFE165&logo=arxiv&style=for-the-badge" alt="Paper on Arxiv"></a>
-  <a href="https://huggingface.co/spaces/OpenHands/evaluation"><img src="https://img.shields.io/badge/Benchmark%20score-000?logoColor=FFE165&logo=huggingface&style=for-the-badge" alt="Evaluation Benchmark Score"></a>
-  <hr>
-</div>
-
-Welcome to OpenHands (formerly OpenDevin), a platform for software development agents powered by AI.
-
-OpenHands agents can do anything a human developer can: modify code, run commands, browse the web,
-call APIs, and yes—even copy code snippets from StackOverflow.
-
-Learn more at [docs.all-hands.dev](https://docs.all-hands.dev), or jump to the [Quick Start](#-quick-start).
-
-> [!IMPORTANT]
-> Using OpenHands for work? We'd love to chat! Fill out
-> [this short form](https://docs.google.com/forms/d/e/1FAIpQLSet3VbGaz8z32gW9Wm-Grl4jpt5WgMXPgJ4EDPVmCETCBpJtQ/viewform)
-> to join our Design Partner program, where you'll get early access to commercial features and the opportunity to provide input on our product roadmap.
-
-![App screenshot](./docs/static/img/screenshot.png)
-
-## ⚡ Quick Start
-
-The easiest way to run OpenHands is in Docker.
-See the [Running OpenHands](https://docs.all-hands.dev/modules/usage/installation) guide for
-system requirements and more information.
-
-```bash
-docker pull docker.all-hands.dev/all-hands-ai/runtime:0.27-nikolaik
-
-docker run -it --rm --pull=always \
-    -e SANDBOX_RUNTIME_CONTAINER_IMAGE=docker.all-hands.dev/all-hands-ai/runtime:0.27-nikolaik \
-    -e LOG_ALL_EVENTS=true \
-    -v /var/run/docker.sock:/var/run/docker.sock \
-    -v ~/.openhands-state:/.openhands-state \
-    -p 3000:3000 \
-    --add-host host.docker.internal:host-gateway \
-    --name openhands-app \
-    docker.all-hands.dev/all-hands-ai/openhands:0.27
-```
-
-You'll find OpenHands running at [http://localhost:3000](http://localhost:3000)!
-
-Finally, you'll need a model provider and API key.
-[Anthropic's Claude 3.5 Sonnet](https://www.anthropic.com/api) (`anthropic/claude-3-5-sonnet-20241022`)
-works best, but you have [many options](https://docs.all-hands.dev/modules/usage/llms).
-
----
-
-You can also [connect OpenHands to your local filesystem](https://docs.all-hands.dev/modules/usage/runtimes#connecting-to-your-filesystem),
-run OpenHands in a scriptable [headless mode](https://docs.all-hands.dev/modules/usage/how-to/headless-mode),
-interact with it via a [friendly CLI](https://docs.all-hands.dev/modules/usage/how-to/cli-mode),
-or run it on tagged issues with [a github action](https://docs.all-hands.dev/modules/usage/how-to/github-action).
-
-Visit [Running OpenHands](https://docs.all-hands.dev/modules/usage/installation) for more information and setup instructions.
-
-> [!CAUTION]
-> OpenHands is meant to be run by a single user on their local workstation.
-> It is not appropriate for multi-tenant deployments where multiple users share the same instance. There is no built-in isolation or scalability.
->
-> If you're interested in running OpenHands in a multi-tenant environment, please
-> [get in touch with us](https://docs.google.com/forms/d/e/1FAIpQLSet3VbGaz8z32gW9Wm-Grl4jpt5WgMXPgJ4EDPVmCETCBpJtQ/viewform)
-> for advanced deployment options.
-
-If you want to modify the OpenHands source code, check out [Development.md](https://github.com/All-Hands-AI/OpenHands/blob/main/Development.md).
-
-Having issues? The [Troubleshooting Guide](https://docs.all-hands.dev/modules/usage/troubleshooting) can help.
-
-## 📖 Documentation
-
-To learn more about the project, and for tips on using OpenHands,
-check out our [documentation](https://docs.all-hands.dev/modules/usage/getting-started).
-
-There you'll find resources on how to use different LLM providers,
-troubleshooting resources, and advanced configuration options.
-
-## 🤝 How to Join the Community
-
-OpenHands is a community-driven project, and we welcome contributions from everyone. We do most of our communication
-through Slack, so this is the best place to start, but we also are happy to have you contact us on Discord or Github:
-
-- [Join our Slack workspace](https://join.slack.com/t/openhands-ai/shared_invite/zt-2ypg5jweb-d~6hObZDbXi_HEL8PDrbHg) - Here we talk about research, architecture, and future development.
-- [Join our Discord server](https://discord.gg/ESHStjSjD4) - This is a community-run server for general discussion, questions, and feedback.
-- [Read or post Github Issues](https://github.com/All-Hands-AI/OpenHands/issues) - Check out the issues we're working on, or add your own ideas.
-
-See more about the community in [COMMUNITY.md](./COMMUNITY.md) or find details on contributing in [CONTRIBUTING.md](./CONTRIBUTING.md).
-
-## 📈 Progress
-
-See the monthly OpenHands roadmap [here](https://github.com/orgs/All-Hands-AI/projects/1) (updated at the maintainer's meeting at the end of each month).
-
-<p align="center">
-  <a href="https://star-history.com/#All-Hands-AI/OpenHands&Date">
-    <img src="https://api.star-history.com/svg?repos=All-Hands-AI/OpenHands&type=Date" width="500" alt="Star History Chart">
-  </a>
-</p>
-
-## 📜 License
-
-Distributed under the MIT License. See [`LICENSE`](./LICENSE) for more information.
-
-## 🙏 Acknowledgements
-
-OpenHands is built by a large number of contributors, and every contribution is greatly appreciated! We also build upon other open source projects, and we are deeply thankful for their work.
-
-For a list of open source projects and licenses used in OpenHands, please see our [CREDITS.md](./CREDITS.md) file.
-
-## 📚 Cite
-
-```
-@misc{openhands,
-      title={{OpenHands: An Open Platform for AI Software Developers as Generalist Agents}},
-      author={Xingyao Wang and Boxuan Li and Yufan Song and Frank F. Xu and Xiangru Tang and Mingchen Zhuge and Jiayi Pan and Yueqi Song and Bowen Li and Jaskirat Singh and Hoang H. Tran and Fuqiang Li and Ren Ma and Mingzhang Zheng and Bill Qian and Yanjun Shao and Niklas Muennighoff and Yizhe Zhang and Binyuan Hui and Junyang Lin and Robert Brennan and Hao Peng and Heng Ji and Graham Neubig},
-      year={2024},
-      eprint={2407.16741},
-      archivePrefix={arXiv},
-      primaryClass={cs.SE},
-      url={https://arxiv.org/abs/2407.16741},
-}
-```

agenthub/__init__.py

@@ -12,15 +12,15 @@ from openhands.agenthub import (  # noqa: E402
     codeact_agent,
     delegator_agent,
     dummy_agent,
-    visualbrowsing_agent,
+    planner_agent,
 )
 
 __all__ = [
     'codeact_agent',
+    'planner_agent',
     'delegator_agent',
     'dummy_agent',
     'browsing_agent',
-    'visualbrowsing_agent',
 ]
 
 for agent in all_microagents.values():

agenthub/browsing_agent/response_parser.py

@@ -10,24 +10,17 @@ from openhands.events.action import (
 
 
 class BrowsingResponseParser(ResponseParser):
-    def __init__(self) -> None:
+    def __init__(self):
         # Need to pay attention to the item order in self.action_parsers
         super().__init__()
         self.action_parsers = [BrowsingActionParserMessage()]
         self.default_parser = BrowsingActionParserBrowseInteractive()
 
-    def parse(
-        self, response: str | dict[str, list[dict[str, dict[str, str | None]]]]
-    ) -> Action:
-        if isinstance(response, str):
-            action_str = response
-        else:
-            action_str = self.parse_response(response)
+    def parse(self, response: str) -> Action:
+        action_str = self.parse_response(response)
         return self.parse_action(action_str)
 
-    def parse_response(
-        self, response: dict[str, list[dict[str, dict[str, str | None]]]]
-    ) -> str:
+    def parse_response(self, response) -> str:
         action_str = response['choices'][0]['message']['content']
         if action_str is None:
             return ''
@@ -54,7 +47,9 @@ class BrowsingActionParserMessage(ActionParser):
     - BrowseInteractiveAction(browser_actions) - unexpected response format, message back to user
     """
 
-    def __init__(self) -> None:
+    def __init__(
+        self,
+    ):
         pass
 
     def check_condition(self, action_str: str) -> bool:
@@ -74,7 +69,9 @@ class BrowsingActionParserBrowseInteractive(ActionParser):
     - BrowseInteractiveAction(browser_actions) - handle send message to user function call in BrowserGym
     """
 
-    def __init__(self) -> None:
+    def __init__(
+        self,
+    ):
         pass
 
     def check_condition(self, action_str: str) -> bool:

agenthub/browsing_agent/utils.py

@@ -5,7 +5,7 @@ from warnings import warn
 import yaml
 
 
-def yaml_parser(message: str) -> tuple[dict, bool, str]:
+def yaml_parser(message):
     """Parse a yaml message for the retry function."""
     # saves gpt-3.5 from some yaml parsing errors
     message = re.sub(r':\s*\n(?=\S|\n)', ': ', message)
@@ -22,9 +22,7 @@ def yaml_parser(message: str) -> tuple[dict, bool, str]:
     return value, valid, retry_message
 
 
-def _compress_chunks(
-    text: str, identifier: str, skip_list: list[str], split_regex: str = '\n\n+'
-) -> tuple[dict[str, str], str]:
+def _compress_chunks(text, identifier, skip_list, split_regex='\n\n+'):
     """Compress a string by replacing redundant chunks by identifiers. Chunks are defined by the split_regex."""
     text_list = re.split(split_regex, text)
     text_list = [chunk.strip() for chunk in text_list]
@@ -46,7 +44,7 @@ def _compress_chunks(
     return def_dict, compressed_text
 
 
-def compress_string(text: str) -> str:
+def compress_string(text):
     """Compress a string by replacing redundant paragraphs and lines with identifiers."""
     # Perform paragraph-level compression
     def_dict, compressed_text = _compress_chunks(
@@ -69,7 +67,7 @@ def compress_string(text: str) -> str:
     return definitions + '\n' + compressed_text
 
 
-def extract_html_tags(text: str, keys: list[str]) -> dict[str, list[str]]:
+def extract_html_tags(text, keys):
     """Extract the content within HTML tags for a list of keys.
 
     Parameters
@@ -104,12 +102,7 @@ class ParseError(Exception):
     pass
 
 
-def parse_html_tags_raise(
-    text: str,
-    keys: list[str] | None = None,
-    optional_keys: list[str] | None = None,
-    merge_multiple: bool = False,
-) -> dict[str, str]:
+def parse_html_tags_raise(text, keys=(), optional_keys=(), merge_multiple=False):
     """A version of parse_html_tags that raises an exception if the parsing is not successful."""
     content_dict, valid, retry_message = parse_html_tags(
         text, keys, optional_keys, merge_multiple=merge_multiple
@@ -119,12 +112,7 @@ def parse_html_tags_raise(
     return content_dict
 
 
-def parse_html_tags(
-    text: str,
-    keys: list[str] | None = None,
-    optional_keys: list[str] | None = None,
-    merge_multiple: bool = False,
-) -> tuple[dict[str, str], bool, str]:
+def parse_html_tags(text, keys=(), optional_keys=(), merge_multiple=False):
     """Satisfy the parse api, extracts 1 match per key and validates that all keys are present
 
     Parameters
@@ -145,12 +133,9 @@ def parse_html_tags(
     str
         A message to be displayed to the agent if the parsing was not successful.
     """
-    keys = keys or []
-    optional_keys = optional_keys or []
-    all_keys = list(keys) + list(optional_keys)
+    all_keys = tuple(keys) + tuple(optional_keys)
     content_dict = extract_html_tags(text, all_keys)
     retry_messages = []
-    result_dict: dict[str, str] = {}
 
     for key in all_keys:
         if key not in content_dict:
@@ -158,6 +143,7 @@ def parse_html_tags(
                 retry_messages.append(f'Missing the key <{key}> in the answer.')
         else:
             val = content_dict[key]
+            content_dict[key] = val[0]
             if len(val) > 1:
                 if not merge_multiple:
                     retry_messages.append(
@@ -165,10 +151,8 @@ def parse_html_tags(
                     )
                 else:
                     # merge the multiple instances
-                    result_dict[key] = '\n'.join(val)
-            else:
-                result_dict[key] = val[0]
+                    content_dict[key] = '\n'.join(val)
 
     valid = len(retry_messages) == 0
     retry_message = '\n'.join(retry_messages)
-    return result_dict, valid, retry_message
+    return content_dict, valid, retry_message

agenthub/codeact_agent/codeact_agent.py

@@ -1,90 +1,444 @@
+import json
+import os
+from collections import deque
+
 from litellm import ModelResponse
 
+import openhands.agenthub.codeact_agent.function_calling as codeact_function_calling
+from openhands.controller.agent import Agent
+from openhands.controller.state.state import State
+from openhands.core.config import AgentConfig
 from openhands.core.logger import openhands_logger as logger
 from openhands.core.message import ImageContent, Message, TextContent
-from openhands.core.schema import ActionType
 from openhands.events.action import (
     Action,
     AgentDelegateAction,
     AgentFinishAction,
-    AgentThinkAction,
     BrowseInteractiveAction,
     BrowseURLAction,
     CmdRunAction,
     FileEditAction,
-    FileReadAction,
     IPythonRunCellAction,
     MessageAction,
 )
-from openhands.events.event import Event
 from openhands.events.observation import (
-    AgentCondensationObservation,
     AgentDelegateObservation,
-    AgentThinkObservation,
     BrowserOutputObservation,
     CmdOutputObservation,
     FileEditObservation,
-    FileReadObservation,
     IPythonRunCellObservation,
     UserRejectObservation,
 )
 from openhands.events.observation.error import ErrorObservation
 from openhands.events.observation.observation import Observation
 from openhands.events.serialization.event import truncate_content
+from openhands.llm.llm import LLM
+from openhands.runtime.plugins import (
+    AgentSkillsRequirement,
+    JupyterRequirement,
+    PluginRequirement,
+)
 from openhands.utils.prompt import PromptManager
 
 
-class ConversationMemory:
-    """Processes event history into a coherent conversation for the agent."""
+class CodeActAgent(Agent):
+    VERSION = '2.2'
+    """
+    The Code Act Agent is a minimalist agent.
+    The agent works by passing the model a list of action-observation pairs and prompting the model to take the next step.
 
-    def __init__(self, prompt_manager: PromptManager):
-        self.prompt_manager = prompt_manager
+    ### Overview
 
-    def process_events(
+    This agent implements the CodeAct idea ([paper](https://arxiv.org/abs/2402.01030), [tweet](https://twitter.com/xingyaow_/status/1754556835703751087)) that consolidates LLM agents’ **act**ions into a unified **code** action space for both *simplicity* and *performance* (see paper for more details).
+
+    The conceptual idea is illustrated below. At each turn, the agent can:
+
+    1. **Converse**: Communicate with humans in natural language to ask for clarification, confirmation, etc.
+    2. **CodeAct**: Choose to perform the task by executing code
+    - Execute any valid Linux `bash` command
+    - Execute any valid `Python` code with [an interactive Python interpreter](https://ipython.org/). This is simulated through `bash` command, see plugin system below for more details.
+
+    ![image](https://github.com/All-Hands-AI/OpenHands/assets/38853559/92b622e3-72ad-4a61-8f41-8c040b6d5fb3)
+
+    """
+
+    sandbox_plugins: list[PluginRequirement] = [
+        # NOTE: AgentSkillsRequirement need to go before JupyterRequirement, since
+        # AgentSkillsRequirement provides a lot of Python functions,
+        # and it needs to be initialized before Jupyter for Jupyter to use those functions.
+        AgentSkillsRequirement(),
+        JupyterRequirement(),
+    ]
+
+    def __init__(
         self,
-        condensed_history: list[Event],
-        initial_messages: list[Message],
-        max_message_chars: int | None = None,
-        vision_is_active: bool = False,
-        enable_som_visual_browsing: bool = False,
-    ) -> list[Message]:
-        """Process state history into a list of messages for the LLM.
+        llm: LLM,
+        config: AgentConfig,
+    ) -> None:
+        """Initializes a new instance of the CodeActAgent class.
 
-        Ensures that tool call actions are processed correctly in function calling mode.
+        Parameters:
+        - llm (LLM): The llm to be used by this agent
+        """
+        super().__init__(llm, config)
+        self.reset()
+
+        self.mock_function_calling = False
+        if not self.llm.is_function_calling_active():
+            logger.info(
+                f'Function calling not enabled for model {self.llm.config.model}. '
+                'Mocking function calling via prompting.'
+            )
+            self.mock_function_calling = True
+
+        # Function calling mode
+        self.tools = codeact_function_calling.get_tools(
+            codeact_enable_browsing=self.config.codeact_enable_browsing,
+            codeact_enable_jupyter=self.config.codeact_enable_jupyter,
+            codeact_enable_llm_editor=self.config.codeact_enable_llm_editor,
+        )
+        logger.debug(
+            f'TOOLS loaded for CodeActAgent: {json.dumps(self.tools, indent=2)}'
+        )
+        self.prompt_manager = PromptManager(
+            microagent_dir=os.path.join(os.path.dirname(__file__), 'micro')
+            if self.config.use_microagents
+            else None,
+            prompt_dir=os.path.join(os.path.dirname(__file__), 'prompts'),
+            disabled_microagents=self.config.disabled_microagents,
+        )
+
+        self.pending_actions: deque[Action] = deque()
+
+    def get_action_message(
+        self,
+        action: Action,
+        pending_tool_call_action_messages: dict[str, Message],
+    ) -> list[Message]:
+        """Converts an action into a message format that can be sent to the LLM.
+
+        This method handles different types of actions and formats them appropriately:
+        1. For tool-based actions (AgentDelegate, CmdRun, IPythonRunCell, FileEdit) and agent-sourced AgentFinish:
+            - In function calling mode: Stores the LLM's response in pending_tool_call_action_messages
+            - In non-function calling mode: Creates a message with the action string
+        2. For MessageActions: Creates a message with the text content and optional image content
 
         Args:
-            state: The state containing the history of events to convert
-            condensed_history: The condensed list of events to process
-            initial_messages: The initial messages to include in the result
-            max_message_chars: The maximum number of characters in the content of an event included
-                in the prompt to the LLM. Larger observations are truncated.
-            vision_is_active: Whether vision is active in the LLM. If True, image URLs will be included.
-            enable_som_visual_browsing: Whether to enable visual browsing for the SOM model.
+            action (Action): The action to convert. Can be one of:
+                - CmdRunAction: For executing bash commands
+                - IPythonRunCellAction: For running IPython code
+                - FileEditAction: For editing files
+                - BrowseInteractiveAction: For browsing the web
+                - AgentFinishAction: For ending the interaction
+                - MessageAction: For sending messages
+            pending_tool_call_action_messages (dict[str, Message]): Dictionary mapping response IDs
+                to their corresponding messages. Used in function calling mode to track tool calls
+                that are waiting for their results.
+
+        Returns:
+            list[Message]: A list containing the formatted message(s) for the action.
+                May be empty if the action is handled as a tool call in function calling mode.
+
+        Note:
+            In function calling mode, tool-based actions are stored in pending_tool_call_action_messages
+            rather than being returned immediately. They will be processed later when all corresponding
+            tool call results are available.
         """
-        events = condensed_history
+        # create a regular message from an event
+        if isinstance(
+            action,
+            (
+                AgentDelegateAction,
+                IPythonRunCellAction,
+                FileEditAction,
+                BrowseInteractiveAction,
+                BrowseURLAction,
+            ),
+        ) or (isinstance(action, CmdRunAction) and action.source == 'agent'):
+            tool_metadata = action.tool_call_metadata
+            assert tool_metadata is not None, (
+                'Tool call metadata should NOT be None when function calling is enabled. Action: '
+                + str(action)
+            )
 
-        # Process special events first (system prompts, etc.)
-        messages = initial_messages
+            llm_response: ModelResponse = tool_metadata.model_response
+            assistant_msg = llm_response.choices[0].message
+
+            # Add the LLM message (assistant) that initiated the tool calls
+            # (overwrites any previous message with the same response_id)
+            logger.debug(
+                f'Tool calls type: {type(assistant_msg.tool_calls)}, value: {assistant_msg.tool_calls}'
+            )
+            pending_tool_call_action_messages[llm_response.id] = Message(
+                role=assistant_msg.role,
+                # tool call content SHOULD BE a string
+                content=[TextContent(text=assistant_msg.content or '')]
+                if assistant_msg.content is not None
+                else [],
+                tool_calls=assistant_msg.tool_calls,
+            )
+            return []
+        elif isinstance(action, AgentFinishAction):
+            role = 'user' if action.source == 'user' else 'assistant'
+
+            # when agent finishes, it has tool_metadata
+            # which has already been executed, and it doesn't have a response
+            # when the user finishes (/exit), we don't have tool_metadata
+            tool_metadata = action.tool_call_metadata
+            if tool_metadata is not None:
+                # take the response message from the tool call
+                assistant_msg = tool_metadata.model_response.choices[0].message
+                content = assistant_msg.content or ''
+
+                # save content if any, to thought
+                if action.thought:
+                    if action.thought != content:
+                        action.thought += '\n' + content
+                else:
+                    action.thought = content
+
+                # remove the tool call metadata
+                action.tool_call_metadata = None
+            return [
+                Message(
+                    role=role,
+                    content=[TextContent(text=action.thought)],
+                )
+            ]
+        elif isinstance(action, MessageAction):
+            role = 'user' if action.source == 'user' else 'assistant'
+            content = [TextContent(text=action.content or '')]
+            if self.llm.vision_is_active() and action.image_urls:
+                content.append(ImageContent(image_urls=action.image_urls))
+            return [
+                Message(
+                    role=role,
+                    content=content,
+                )
+            ]
+        elif isinstance(action, CmdRunAction) and action.source == 'user':
+            content = [
+                TextContent(text=f'User executed the command:\n{action.command}')
+            ]
+            return [
+                Message(
+                    role='user',
+                    content=content,
+                )
+            ]
+        return []
+
+    def get_observation_message(
+        self,
+        obs: Observation,
+        tool_call_id_to_message: dict[str, Message],
+    ) -> list[Message]:
+        """Converts an observation into a message format that can be sent to the LLM.
+
+        This method handles different types of observations and formats them appropriately:
+        - CmdOutputObservation: Formats command execution results with exit codes
+        - IPythonRunCellObservation: Formats IPython cell execution results, replacing base64 images
+        - FileEditObservation: Formats file editing results
+        - AgentDelegateObservation: Formats results from delegated agent tasks
+        - ErrorObservation: Formats error messages from failed actions
+        - UserRejectObservation: Formats user rejection messages
+
+        In function calling mode, observations with tool_call_metadata are stored in
+        tool_call_id_to_message for later processing instead of being returned immediately.
+
+        Args:
+            obs (Observation): The observation to convert
+            tool_call_id_to_message (dict[str, Message]): Dictionary mapping tool call IDs
+                to their corresponding messages (used in function calling mode)
+
+        Returns:
+            list[Message]: A list containing the formatted message(s) for the observation.
+                May be empty if the observation is handled as a tool response in function calling mode.
+
+        Raises:
+            ValueError: If the observation type is unknown
+        """
+        message: Message
+        max_message_chars = self.llm.config.max_message_chars
+        if isinstance(obs, CmdOutputObservation):
+            # if it doesn't have tool call metadata, it was triggered by a user action
+            if obs.tool_call_metadata is None:
+                text = truncate_content(
+                    f'\nObserved result of command executed by user:\n{obs.content}',
+                    max_message_chars,
+                )
+            else:
+                text = truncate_content(
+                    obs.content + obs.interpreter_details, max_message_chars
+                )
+            text += f'\n[Command finished with exit code {obs.exit_code}]'
+            message = Message(role='user', content=[TextContent(text=text)])
+        elif isinstance(obs, IPythonRunCellObservation):
+            text = obs.content
+            # replace base64 images with a placeholder
+            splitted = text.split('\n')
+            for i, line in enumerate(splitted):
+                if '![image](data:image/png;base64,' in line:
+                    splitted[i] = (
+                        '![image](data:image/png;base64, ...) already displayed to user'
+                    )
+            text = '\n'.join(splitted)
+            text = truncate_content(text, max_message_chars)
+            message = Message(role='user', content=[TextContent(text=text)])
+        elif isinstance(obs, FileEditObservation):
+            text = truncate_content(str(obs), max_message_chars)
+            message = Message(role='user', content=[TextContent(text=text)])
+        elif isinstance(obs, BrowserOutputObservation):
+            text = obs.get_agent_obs_text()
+            message = Message(
+                role='user',
+                content=[TextContent(text=text)],
+            )
+        elif isinstance(obs, AgentDelegateObservation):
+            text = truncate_content(
+                obs.outputs['content'] if 'content' in obs.outputs else '',
+                max_message_chars,
+            )
+            message = Message(role='user', content=[TextContent(text=text)])
+        elif isinstance(obs, ErrorObservation):
+            text = truncate_content(obs.content, max_message_chars)
+            text += '\n[Error occurred in processing last action]'
+            message = Message(role='user', content=[TextContent(text=text)])
+        elif isinstance(obs, UserRejectObservation):
+            text = 'OBSERVATION:\n' + truncate_content(obs.content, max_message_chars)
+            text += '\n[Last action has been rejected by the user]'
+            message = Message(role='user', content=[TextContent(text=text)])
+        else:
+            # If an observation message is not returned, it will cause an error
+            # when the LLM tries to return the next message
+            raise ValueError(f'Unknown observation type: {type(obs)}')
+
+        # Update the message as tool response properly
+        if (tool_call_metadata := obs.tool_call_metadata) is not None:
+            tool_call_id_to_message[tool_call_metadata.tool_call_id] = Message(
+                role='tool',
+                content=message.content,
+                tool_call_id=tool_call_metadata.tool_call_id,
+                name=tool_call_metadata.function_name,
+            )
+            # No need to return the observation message
+            # because it will be added by get_action_message when all the corresponding
+            # tool calls in the SAME request are processed
+            return []
+
+        return [message]
+
+    def reset(self) -> None:
+        """Resets the CodeAct Agent."""
+        super().reset()
+
+    def step(self, state: State) -> Action:
+        """Performs one step using the CodeAct Agent.
+        This includes gathering info on previous steps and prompting the model to make a command to execute.
+
+        Parameters:
+        - state (State): used to get updated info
+
+        Returns:
+        - CmdRunAction(command) - bash command to run
+        - IPythonRunCellAction(code) - IPython code to run
+        - AgentDelegateAction(agent, inputs) - delegate action for (sub)task
+        - MessageAction(content) - Message action to run (e.g. ask for clarification)
+        - AgentFinishAction() - end the interaction
+        """
+        # Continue with pending actions if any
+        if self.pending_actions:
+            return self.pending_actions.popleft()
+
+        # if we're done, go back
+        latest_user_message = state.get_last_user_message()
+        if latest_user_message and latest_user_message.content.strip() == '/exit':
+            return AgentFinishAction()
+
+        # prepare what we want to send to the LLM
+        messages = self._get_messages(state)
+        params: dict = {
+            'messages': self.llm.format_messages_for_llm(messages),
+        }
+        params['tools'] = self.tools
+        if self.mock_function_calling:
+            params['mock_function_calling'] = True
+        response = self.llm.completion(**params)
+        actions = codeact_function_calling.response_to_actions(response)
+        for action in actions:
+            self.pending_actions.append(action)
+        return self.pending_actions.popleft()
+
+    def _get_messages(self, state: State) -> list[Message]:
+        """Constructs the message history for the LLM conversation.
+
+        This method builds a structured conversation history by processing events from the state
+        and formatting them into messages that the LLM can understand. It handles both regular
+        message flow and function-calling scenarios.
+
+        The method performs the following steps:
+        1. Initializes with system prompt and optional initial user message
+        2. Processes events (Actions and Observations) into messages
+        3. Handles tool calls and their responses in function-calling mode
+        4. Manages message role alternation (user/assistant/tool)
+        5. Applies caching for specific LLM providers (e.g., Anthropic)
+        6. Adds environment reminders for non-function-calling mode
+
+        Args:
+            state (State): The current state object containing conversation history and other metadata
+
+        Returns:
+            list[Message]: A list of formatted messages ready for LLM consumption, including:
+                - System message with prompt
+                - Initial user message (if configured)
+                - Action messages (from both user and assistant)
+                - Observation messages (including tool responses)
+                - Environment reminders (in non-function-calling mode)
+
+        Note:
+            - In function-calling mode, tool calls and their responses are carefully tracked
+              to maintain proper conversation flow
+            - Messages from the same role are combined to prevent consecutive same-role messages
+            - For Anthropic models, specific messages are cached according to their documentation
+        """
+        if not self.prompt_manager:
+            raise Exception('Prompt Manager not instantiated.')
+
+        messages: list[Message] = [
+            Message(
+                role='system',
+                content=[
+                    TextContent(
+                        text=self.prompt_manager.get_system_message(),
+                        cache_prompt=self.llm.is_caching_prompt_active(),
+                    )
+                ],
+            )
+        ]
+        example_message = self.prompt_manager.get_example_user_message()
+        if example_message:
+            messages.append(
+                Message(
+                    role='user',
+                    content=[TextContent(text=example_message)],
+                    cache_prompt=self.llm.is_caching_prompt_active(),
+                )
+            )
 
-        # Process regular events
         pending_tool_call_action_messages: dict[str, Message] = {}
         tool_call_id_to_message: dict[str, Message] = {}
-
+        events = list(state.history)
         for event in events:
             # create a regular message from an event
             if isinstance(event, Action):
-                messages_to_add = self._process_action(
+                messages_to_add = self.get_action_message(
                     action=event,
                     pending_tool_call_action_messages=pending_tool_call_action_messages,
-                    vision_is_active=vision_is_active,
                 )
             elif isinstance(event, Observation):
-                messages_to_add = self._process_observation(
+                messages_to_add = self.get_observation_message(
                     obs=event,
                     tool_call_id_to_message=tool_call_id_to_message,
-                    max_message_chars=max_message_chars,
-                    vision_is_active=vision_is_active,
-                    enable_som_visual_browsing=enable_som_visual_browsing,
                 )
             else:
                 raise ValueError(f'Unknown event type: {type(event)}')
@@ -115,292 +469,36 @@ class ConversationMemory:
             for response_id in _response_ids_to_remove:
                 pending_tool_call_action_messages.pop(response_id)
 
-            messages += messages_to_add
+            for message in messages_to_add:
+                if message:
+                    if message.role == 'user':
+                        self.prompt_manager.enhance_message(message)
+                    # handle error if the message is the SAME role as the previous message
+                    # litellm.exceptions.BadRequestError: litellm.BadRequestError: OpenAIException - Error code: 400 - {'detail': 'Only supports u/a/u/a/u...'}
+                    # there shouldn't be two consecutive messages from the same role
+                    # NOTE: we shouldn't combine tool messages because each of them has a different tool_call_id
+                    if (
+                        messages
+                        and messages[-1].role == message.role
+                        and message.role != 'tool'
+                    ):
+                        messages[-1].content.extend(message.content)
+                    else:
+                        messages.append(message)
+
+        if self.llm.is_caching_prompt_active():
+            # NOTE: this is only needed for anthropic
+            # following logic here:
+            # https://github.com/anthropics/anthropic-quickstarts/blob/8f734fd08c425c6ec91ddd613af04ff87d70c5a0/computer-use-demo/computer_use_demo/loop.py#L241-L262
+            breakpoints_remaining = 3  # remaining 1 for system/tool
+            for message in reversed(messages):
+                if message.role == 'user' or message.role == 'tool':
+                    if breakpoints_remaining > 0:
+                        message.content[
+                            -1
+                        ].cache_prompt = True  # Last item inside the message content
+                        breakpoints_remaining -= 1
+                    else:
+                        break
 
         return messages
-
-    def process_initial_messages(self, with_caching: bool = False) -> list[Message]:
-        """Create the initial messages for the conversation."""
-        return [
-            Message(
-                role='system',
-                content=[
-                    TextContent(
-                        text=self.prompt_manager.get_system_message(),
-                        cache_prompt=with_caching,
-                    )
-                ],
-            )
-        ]
-
-    def _process_action(
-        self,
-        action: Action,
-        pending_tool_call_action_messages: dict[str, Message],
-        vision_is_active: bool = False,
-    ) -> list[Message]:
-        """Converts an action into a message format that can be sent to the LLM.
-
-        This method handles different types of actions and formats them appropriately:
-        1. For tool-based actions (AgentDelegate, CmdRun, IPythonRunCell, FileEdit) and agent-sourced AgentFinish:
-            - In function calling mode: Stores the LLM's response in pending_tool_call_action_messages
-            - In non-function calling mode: Creates a message with the action string
-        2. For MessageActions: Creates a message with the text content and optional image content
-
-        Args:
-            action: The action to convert. Can be one of:
-                - CmdRunAction: For executing bash commands
-                - IPythonRunCellAction: For running IPython code
-                - FileEditAction: For editing files
-                - FileReadAction: For reading files using openhands-aci commands
-                - BrowseInteractiveAction: For browsing the web
-                - AgentFinishAction: For ending the interaction
-                - MessageAction: For sending messages
-
-            pending_tool_call_action_messages: Dictionary mapping response IDs to their corresponding messages.
-                Used in function calling mode to track tool calls that are waiting for their results.
-
-            vision_is_active: Whether vision is active in the LLM. If True, image URLs will be included
-
-        Returns:
-            list[Message]: A list containing the formatted message(s) for the action.
-                May be empty if the action is handled as a tool call in function calling mode.
-
-        Note:
-            In function calling mode, tool-based actions are stored in pending_tool_call_action_messages
-            rather than being returned immediately. They will be processed later when all corresponding
-            tool call results are available.
-        """
-        # create a regular message from an event
-        if isinstance(
-            action,
-            (
-                AgentDelegateAction,
-                AgentThinkAction,
-                IPythonRunCellAction,
-                FileEditAction,
-                FileReadAction,
-                BrowseInteractiveAction,
-                BrowseURLAction,
-            ),
-        ) or (isinstance(action, CmdRunAction) and action.source == 'agent'):
-            tool_metadata = action.tool_call_metadata
-            assert tool_metadata is not None, (
-                'Tool call metadata should NOT be None when function calling is enabled. Action: '
-                + str(action)
-            )
-
-            llm_response: ModelResponse = tool_metadata.model_response
-            assistant_msg = getattr(llm_response.choices[0], 'message')
-
-            # Add the LLM message (assistant) that initiated the tool calls
-            # (overwrites any previous message with the same response_id)
-            logger.debug(
-                f'Tool calls type: {type(assistant_msg.tool_calls)}, value: {assistant_msg.tool_calls}'
-            )
-            pending_tool_call_action_messages[llm_response.id] = Message(
-                role=getattr(assistant_msg, 'role', 'assistant'),
-                # tool call content SHOULD BE a string
-                content=[TextContent(text=assistant_msg.content or '')]
-                if assistant_msg.content is not None
-                else [],
-                tool_calls=assistant_msg.tool_calls,
-            )
-            return []
-        elif isinstance(action, AgentFinishAction):
-            role = 'user' if action.source == 'user' else 'assistant'
-
-            # when agent finishes, it has tool_metadata
-            # which has already been executed, and it doesn't have a response
-            # when the user finishes (/exit), we don't have tool_metadata
-            tool_metadata = action.tool_call_metadata
-            if tool_metadata is not None:
-                # take the response message from the tool call
-                assistant_msg = getattr(
-                    tool_metadata.model_response.choices[0], 'message'
-                )
-                content = assistant_msg.content or ''
-
-                # save content if any, to thought
-                if action.thought:
-                    if action.thought != content:
-                        action.thought += '\n' + content
-                else:
-                    action.thought = content
-
-                # remove the tool call metadata
-                action.tool_call_metadata = None
-            if role not in ('user', 'system', 'assistant', 'tool'):
-                raise ValueError(f'Invalid role: {role}')
-            return [
-                Message(
-                    role=role,  # type: ignore[arg-type]
-                    content=[TextContent(text=action.thought)],
-                )
-            ]
-        elif isinstance(action, MessageAction):
-            role = 'user' if action.source == 'user' else 'assistant'
-            content = [TextContent(text=action.content or '')]
-            if vision_is_active and action.image_urls:
-                content.append(ImageContent(image_urls=action.image_urls))
-            if role not in ('user', 'system', 'assistant', 'tool'):
-                raise ValueError(f'Invalid role: {role}')
-            return [
-                Message(
-                    role=role,  # type: ignore[arg-type]
-                    content=content,
-                )
-            ]
-        elif isinstance(action, CmdRunAction) and action.source == 'user':
-            content = [
-                TextContent(text=f'User executed the command:\n{action.command}')
-            ]
-            return [
-                Message(
-                    role='user',  # Always user for CmdRunAction
-                    content=content,
-                )
-            ]
-        return []
-
-    def _process_observation(
-        self,
-        obs: Observation,
-        tool_call_id_to_message: dict[str, Message],
-        max_message_chars: int | None = None,
-        vision_is_active: bool = False,
-        enable_som_visual_browsing: bool = False,
-    ) -> list[Message]:
-        """Converts an observation into a message format that can be sent to the LLM.
-
-        This method handles different types of observations and formats them appropriately:
-        - CmdOutputObservation: Formats command execution results with exit codes
-        - IPythonRunCellObservation: Formats IPython cell execution results, replacing base64 images
-        - FileEditObservation: Formats file editing results
-        - FileReadObservation: Formats file reading results from openhands-aci
-        - AgentDelegateObservation: Formats results from delegated agent tasks
-        - ErrorObservation: Formats error messages from failed actions
-        - UserRejectObservation: Formats user rejection messages
-
-        In function calling mode, observations with tool_call_metadata are stored in
-        tool_call_id_to_message for later processing instead of being returned immediately.
-
-        Args:
-            obs: The observation to convert
-            tool_call_id_to_message: Dictionary mapping tool call IDs to their corresponding messages (used in function calling mode)
-            max_message_chars: The maximum number of characters in the content of an observation included in the prompt to the LLM
-            vision_is_active: Whether vision is active in the LLM. If True, image URLs will be included
-            enable_som_visual_browsing: Whether to enable visual browsing for the SOM model
-
-        Returns:
-            list[Message]: A list containing the formatted message(s) for the observation.
-                May be empty if the observation is handled as a tool response in function calling mode.
-
-        Raises:
-            ValueError: If the observation type is unknown
-        """
-        message: Message
-
-        if isinstance(obs, CmdOutputObservation):
-            # if it doesn't have tool call metadata, it was triggered by a user action
-            if obs.tool_call_metadata is None:
-                text = truncate_content(
-                    f'\nObserved result of command executed by user:\n{obs.to_agent_observation()}',
-                    max_message_chars,
-                )
-            else:
-                text = truncate_content(obs.to_agent_observation(), max_message_chars)
-            message = Message(role='user', content=[TextContent(text=text)])
-        elif isinstance(obs, IPythonRunCellObservation):
-            text = obs.content
-            # replace base64 images with a placeholder
-            splitted = text.split('\n')
-            for i, line in enumerate(splitted):
-                if '![image](data:image/png;base64,' in line:
-                    splitted[i] = (
-                        '![image](data:image/png;base64, ...) already displayed to user'
-                    )
-            text = '\n'.join(splitted)
-            text = truncate_content(text, max_message_chars)
-            message = Message(role='user', content=[TextContent(text=text)])
-        elif isinstance(obs, FileEditObservation):
-            text = truncate_content(str(obs), max_message_chars)
-            message = Message(role='user', content=[TextContent(text=text)])
-        elif isinstance(obs, FileReadObservation):
-            message = Message(
-                role='user', content=[TextContent(text=obs.content)]
-            )  # Content is already truncated by openhands-aci
-        elif isinstance(obs, BrowserOutputObservation):
-            text = obs.get_agent_obs_text()
-            if (
-                obs.trigger_by_action == ActionType.BROWSE_INTERACTIVE
-                and obs.set_of_marks is not None
-                and len(obs.set_of_marks) > 0
-                and enable_som_visual_browsing
-                and vision_is_active
-            ):
-                text += 'Image: Current webpage screenshot (Note that only visible portion of webpage is present in the screenshot. You may need to scroll to view the remaining portion of the web-page.)\n'
-                message = Message(
-                    role='user',
-                    content=[
-                        TextContent(text=text),
-                        ImageContent(image_urls=[obs.set_of_marks]),
-                    ],
-                )
-            else:
-                message = Message(
-                    role='user',
-                    content=[TextContent(text=text)],
-                )
-        elif isinstance(obs, AgentDelegateObservation):
-            text = truncate_content(
-                obs.outputs['content'] if 'content' in obs.outputs else '',
-                max_message_chars,
-            )
-            message = Message(role='user', content=[TextContent(text=text)])
-        elif isinstance(obs, AgentThinkObservation):
-            text = truncate_content(obs.content, max_message_chars)
-            message = Message(role='user', content=[TextContent(text=text)])
-        elif isinstance(obs, ErrorObservation):
-            text = truncate_content(obs.content, max_message_chars)
-            text += '\n[Error occurred in processing last action]'
-            message = Message(role='user', content=[TextContent(text=text)])
-        elif isinstance(obs, UserRejectObservation):
-            text = 'OBSERVATION:\n' + truncate_content(obs.content, max_message_chars)
-            text += '\n[Last action has been rejected by the user]'
-            message = Message(role='user', content=[TextContent(text=text)])
-        elif isinstance(obs, AgentCondensationObservation):
-            text = truncate_content(obs.content, max_message_chars)
-            message = Message(role='user', content=[TextContent(text=text)])
-        else:
-            # If an observation message is not returned, it will cause an error
-            # when the LLM tries to return the next message
-            raise ValueError(f'Unknown observation type: {type(obs)}')
-
-        # Update the message as tool response properly
-        if (tool_call_metadata := getattr(obs, 'tool_call_metadata', None)) is not None:
-            tool_call_id_to_message[tool_call_metadata.tool_call_id] = Message(
-                role='tool',
-                content=message.content,
-                tool_call_id=tool_call_metadata.tool_call_id,
-                name=tool_call_metadata.function_name,
-            )
-            # No need to return the observation message
-            # because it will be added by get_action_message when all the corresponding
-            # tool calls in the SAME request are processed
-            return []
-
-        return [message]
-
-    def apply_prompt_caching(self, messages: list[Message]) -> None:
-        """Applies caching breakpoints to the messages.
-
-        For new Anthropic API, we only need to mark the last user or tool message as cacheable.
-        """
-        # NOTE: this is only needed for anthropic
-        for message in reversed(messages):
-            if message.role in ('user', 'tool'):
-                message.content[
-                    -1
-                ].cache_prompt = True  # Last item inside the message content
-                break

agenthub/codeact_agent/function_calling.py

@@ -0,0 +1,554 @@
+"""This file contains the function calling implementation for different actions.
+
+This is similar to the functionality of `CodeActResponseParser`.
+"""
+
+import json
+
+from browsergym.core.action.highlevel import HighLevelActionSet
+from litellm import (
+    ChatCompletionToolParam,
+    ChatCompletionToolParamFunctionChunk,
+    ModelResponse,
+)
+
+from openhands.core.exceptions import FunctionCallNotExistsError
+from openhands.core.logger import openhands_logger as logger
+from openhands.events.action import (
+    Action,
+    AgentDelegateAction,
+    AgentFinishAction,
+    BrowseInteractiveAction,
+    BrowseURLAction,
+    CmdRunAction,
+    FileEditAction,
+    IPythonRunCellAction,
+    MessageAction,
+)
+from openhands.events.tool import ToolCallMetadata
+
+_BASH_DESCRIPTION = """Execute a bash command in the terminal.
+* Long running commands: For commands that may run indefinitely, it should be run in the background and the output should be redirected to a file, e.g. command = `python3 app.py > server.log 2>&1 &`.
+* Interactive: If a bash command returns exit code `-1`, this means the process is not yet finished. The assistant must then send a second call to terminal with an empty `command` (which will retrieve any additional logs), or it can send additional text (set `command` to the text) to STDIN of the running process, or it can send command=`ctrl+c` to interrupt the process.
+* Timeout: If a command execution result says "Command timed out. Sending SIGINT to the process", the assistant should retry running the command in the background.
+"""
+
+CmdRunTool = ChatCompletionToolParam(
+    type='function',
+    function=ChatCompletionToolParamFunctionChunk(
+        name='execute_bash',
+        description=_BASH_DESCRIPTION,
+        parameters={
+            'type': 'object',
+            'properties': {
+                'command': {
+                    'type': 'string',
+                    'description': 'The bash command to execute. Can be empty to view additional logs when previous exit code is `-1`. Can be `ctrl+c` to interrupt the currently running process.',
+                },
+            },
+            'required': ['command'],
+        },
+    ),
+)
+
+_IPYTHON_DESCRIPTION = """Run a cell of Python code in an IPython environment.
+* The assistant should define variables and import packages before using them.
+* The variable defined in the IPython environment will not be available outside the IPython environment (e.g., in terminal).
+"""
+
+IPythonTool = ChatCompletionToolParam(
+    type='function',
+    function=ChatCompletionToolParamFunctionChunk(
+        name='execute_ipython_cell',
+        description=_IPYTHON_DESCRIPTION,
+        parameters={
+            'type': 'object',
+            'properties': {
+                'code': {
+                    'type': 'string',
+                    'description': 'The Python code to execute. Supports magic commands like %pip.',
+                },
+            },
+            'required': ['code'],
+        },
+    ),
+)
+
+_FILE_EDIT_DESCRIPTION = """Edit a file.
+* The assistant can edit files by specifying the file path and providing a draft of the new file content.
+* The draft content doesn't need to be exactly the same as the existing file; the assistant may skip unchanged lines using comments like `# unchanged` to indicate unchanged sections.
+* IMPORTANT: For large files (e.g., > 300 lines), specify the range of lines to edit using `start` and `end` (1-indexed, inclusive). The range should be smaller than 300 lines.
+* To append to a file, set both `start` and `end` to `-1`.
+* If the file doesn't exist, a new file will be created with the provided content.
+
+**Example 1: general edit for short files**
+For example, given an existing file `/path/to/file.py` that looks like this:
+(this is the end of the file)
+1|class MyClass:
+2|    def __init__(self):
+3|        self.x = 1
+4|        self.y = 2
+5|        self.z = 3
+6|
+7|print(MyClass().z)
+8|print(MyClass().x)
+(this is the end of the file)
+
+The assistant wants to edit the file to look like this:
+(this is the end of the file)
+1|class MyClass:
+2|    def __init__(self):
+3|        self.x = 1
+4|        self.y = 2
+5|
+6|print(MyClass().y)
+(this is the end of the file)
+
+The assistant may produce an edit action like this:
+path="/path/to/file.txt" start=1 end=-1
+content=```
+class MyClass:
+    def __init__(self):
+        # no changes before
+        self.y = 2
+        # self.z is removed
+
+# MyClass().z is removed
+print(MyClass().y)
+```
+
+**Example 2: append to file for short files**
+For example, given an existing file `/path/to/file.py` that looks like this:
+(this is the end of the file)
+1|class MyClass:
+2|    def __init__(self):
+3|        self.x = 1
+4|        self.y = 2
+5|        self.z = 3
+6|
+7|print(MyClass().z)
+8|print(MyClass().x)
+(this is the end of the file)
+
+To append the following lines to the file:
+```python
+print(MyClass().y)
+```
+
+The assistant may produce an edit action like this:
+path="/path/to/file.txt" start=-1 end=-1
+content=```
+print(MyClass().y)
+```
+
+**Example 3: edit for long files**
+
+Given an existing file `/path/to/file.py` that looks like this:
+(1000 more lines above)
+1001|class MyClass:
+1002|    def __init__(self):
+1003|        self.x = 1
+1004|        self.y = 2
+1005|        self.z = 3
+1006|
+1007|print(MyClass().z)
+1008|print(MyClass().x)
+(2000 more lines below)
+
+The assistant wants to edit the file to look like this:
+
+(1000 more lines above)
+1001|class MyClass:
+1002|    def __init__(self):
+1003|        self.x = 1
+1004|        self.y = 2
+1005|
+1006|print(MyClass().y)
+(2000 more lines below)
+
+The assistant may produce an edit action like this:
+path="/path/to/file.txt" start=1001 end=1008
+content=```
+class MyClass:
+    def __init__(self):
+        # no changes before
+        self.y = 2
+        # self.z is removed
+
+# MyClass().z is removed
+print(MyClass().y)
+```
+"""
+
+LLMBasedFileEditTool = ChatCompletionToolParam(
+    type='function',
+    function=ChatCompletionToolParamFunctionChunk(
+        name='edit_file',
+        description=_FILE_EDIT_DESCRIPTION,
+        parameters={
+            'type': 'object',
+            'properties': {
+                'path': {
+                    'type': 'string',
+                    'description': 'The absolute path to the file to be edited.',
+                },
+                'new_content_draft': {
+                    'type': 'string',
+                    'description': 'A draft of the new content for the file being edited. Note that the assistant may skip unchanged lines.',
+                },
+                'start': {
+                    'type': 'integer',
+                    'description': 'The starting line number for the edit (1-indexed, inclusive). Default is 1.',
+                },
+                'end': {
+                    'type': 'integer',
+                    'description': 'The ending line number for the edit (1-indexed, inclusive). Default is -1 (end of file).',
+                },
+            },
+            'required': ['path', 'content'],
+        },
+    ),
+)
+
+_STR_REPLACE_EDITOR_DESCRIPTION = """Custom editing tool for viewing, creating and editing files
+* State is persistent across command calls and discussions with the user
+* If `path` is a file, `view` displays the result of applying `cat -n`. If `path` is a directory, `view` lists non-hidden files and directories up to 2 levels deep
+* The `create` command cannot be used if the specified `path` already exists as a file
+* If a `command` generates a long output, it will be truncated and marked with `<response clipped>`
+* The `undo_edit` command will revert the last edit made to the file at `path`
+
+Notes for using the `str_replace` command:
+* The `old_str` parameter should match EXACTLY one or more consecutive lines from the original file. Be mindful of whitespaces!
+* If the `old_str` parameter is not unique in the file, the replacement will not be performed. Make sure to include enough context in `old_str` to make it unique
+* The `new_str` parameter should contain the edited lines that should replace the `old_str`
+"""
+
+StrReplaceEditorTool = ChatCompletionToolParam(
+    type='function',
+    function=ChatCompletionToolParamFunctionChunk(
+        name='str_replace_editor',
+        description=_STR_REPLACE_EDITOR_DESCRIPTION,
+        parameters={
+            'type': 'object',
+            'properties': {
+                'command': {
+                    'description': 'The commands to run. Allowed options are: `view`, `create`, `str_replace`, `insert`, `undo_edit`.',
+                    'enum': ['view', 'create', 'str_replace', 'insert', 'undo_edit'],
+                    'type': 'string',
+                },
+                'path': {
+                    'description': 'Absolute path to file or directory, e.g. `/workspace/file.py` or `/workspace`.',
+                    'type': 'string',
+                },
+                'file_text': {
+                    'description': 'Required parameter of `create` command, with the content of the file to be created.',
+                    'type': 'string',
+                },
+                'old_str': {
+                    'description': 'Required parameter of `str_replace` command containing the string in `path` to replace.',
+                    'type': 'string',
+                },
+                'new_str': {
+                    'description': 'Optional parameter of `str_replace` command containing the new string (if not given, no string will be added). Required parameter of `insert` command containing the string to insert.',
+                    'type': 'string',
+                },
+                'insert_line': {
+                    'description': 'Required parameter of `insert` command. The `new_str` will be inserted AFTER the line `insert_line` of `path`.',
+                    'type': 'integer',
+                },
+                'view_range': {
+                    'description': 'Optional parameter of `view` command when `path` points to a file. If none is given, the full file is shown. If provided, the file will be shown in the indicated line number range, e.g. [11, 12] will show lines 11 and 12. Indexing at 1 to start. Setting `[start_line, -1]` shows all lines from `start_line` to the end of the file.',
+                    'items': {'type': 'integer'},
+                    'type': 'array',
+                },
+            },
+            'required': ['command', 'path'],
+        },
+    ),
+)
+
+
+_WEB_DESCRIPTION = """Read (convert to markdown) content from a webpage. You should prefer using the `webpage_read` tool over the `browser` tool, but do use the `browser` tool if you need to interact with a webpage (e.g., click a button, fill out a form, etc.).
+
+You may use the `webpage_read` tool to read content from a webpage, and even search the webpage content using a Google search query (e.g., url=`https://www.google.com/search?q=YOUR_QUERY`).
+"""
+
+WebReadTool = ChatCompletionToolParam(
+    type='function',
+    function=ChatCompletionToolParamFunctionChunk(
+        name='web_read',
+        description=_WEB_DESCRIPTION,
+        parameters={
+            'type': 'object',
+            'properties': {
+                'url': {
+                    'type': 'string',
+                    'description': 'The URL of the webpage to read. You can also use a Google search query here (e.g., `https://www.google.com/search?q=YOUR_QUERY`).',
+                }
+            },
+            'required': ['url'],
+        },
+    ),
+)
+
+# from browsergym/core/action/highlevel.py
+_browser_action_space = HighLevelActionSet(
+    subsets=['bid', 'nav'],
+    strict=False,  # less strict on the parsing of the actions
+    multiaction=True,  # enable to agent to take multiple actions at once
+)
+
+
+_BROWSER_DESCRIPTION = """Interact with the browser using Python code. Use it ONLY when you need to interact with a webpage.
+
+See the description of "code" parameter for more details.
+
+Multiple actions can be provided at once, but will be executed sequentially without any feedback from the page.
+More than 2-3 actions usually leads to failure or unexpected behavior. Example:
+fill('a12', 'example with "quotes"')
+click('a51')
+click('48', button='middle', modifiers=['Shift'])
+"""
+
+_BROWSER_TOOL_DESCRIPTION = """
+The following 15 functions are available. Nothing else is supported.
+
+goto(url: str)
+    Description: Navigate to a url.
+    Examples:
+        goto('http://www.example.com')
+
+go_back()
+    Description: Navigate to the previous page in history.
+    Examples:
+        go_back()
+
+go_forward()
+    Description: Navigate to the next page in history.
+    Examples:
+        go_forward()
+
+noop(wait_ms: float = 1000)
+    Description: Do nothing, and optionally wait for the given time (in milliseconds).
+    You can use this to get the current page content and/or wait for the page to load.
+    Examples:
+        noop()
+
+        noop(500)
+
+scroll(delta_x: float, delta_y: float)
+    Description: Scroll horizontally and vertically. Amounts in pixels, positive for right or down scrolling, negative for left or up scrolling. Dispatches a wheel event.
+    Examples:
+        scroll(0, 200)
+
+        scroll(-50.2, -100.5)
+
+fill(bid: str, value: str)
+    Description: Fill out a form field. It focuses the element and triggers an input event with the entered text. It works for <input>, <textarea> and [contenteditable] elements.
+    Examples:
+        fill('237', 'example value')
+
+        fill('45', 'multi-line\nexample')
+
+        fill('a12', 'example with "quotes"')
+
+select_option(bid: str, options: str | list[str])
+    Description: Select one or multiple options in a <select> element. You can specify option value or label to select. Multiple options can be selected.
+    Examples:
+        select_option('a48', 'blue')
+
+        select_option('c48', ['red', 'green', 'blue'])
+
+click(bid: str, button: Literal['left', 'middle', 'right'] = 'left', modifiers: list[typing.Literal['Alt', 'Control', 'ControlOrMeta', 'Meta', 'Shift']] = [])
+    Description: Click an element.
+    Examples:
+        click('a51')
+
+        click('b22', button='right')
+
+        click('48', button='middle', modifiers=['Shift'])
+
+dblclick(bid: str, button: Literal['left', 'middle', 'right'] = 'left', modifiers: list[typing.Literal['Alt', 'Control', 'ControlOrMeta', 'Meta', 'Shift']] = [])
+    Description: Double click an element.
+    Examples:
+        dblclick('12')
+
+        dblclick('ca42', button='right')
+
+        dblclick('178', button='middle', modifiers=['Shift'])
+
+hover(bid: str)
+    Description: Hover over an element.
+    Examples:
+        hover('b8')
+
+press(bid: str, key_comb: str)
+    Description: Focus the matching element and press a combination of keys. It accepts the logical key names that are emitted in the keyboardEvent.key property of the keyboard events: Backquote, Minus, Equal, Backslash, Backspace, Tab, Delete, Escape, ArrowDown, End, Enter, Home, Insert, PageDown, PageUp, ArrowRight, ArrowUp, F1 - F12, Digit0 - Digit9, KeyA - KeyZ, etc. You can alternatively specify a single character you'd like to produce such as "a" or "#". Following modification shortcuts are also supported: Shift, Control, Alt, Meta, ShiftLeft, ControlOrMeta. ControlOrMeta resolves to Control on Windows and Linux and to Meta on macOS.
+    Examples:
+        press('88', 'Backspace')
+
+        press('a26', 'ControlOrMeta+a')
+
+        press('a61', 'Meta+Shift+t')
+
+focus(bid: str)
+    Description: Focus the matching element.
+    Examples:
+        focus('b455')
+
+clear(bid: str)
+    Description: Clear the input field.
+    Examples:
+        clear('996')
+
+drag_and_drop(from_bid: str, to_bid: str)
+    Description: Perform a drag & drop. Hover the element that will be dragged. Press left mouse button. Move mouse to the element that will receive the drop. Release left mouse button.
+    Examples:
+        drag_and_drop('56', '498')
+
+upload_file(bid: str, file: str | list[str])
+    Description: Click an element and wait for a "filechooser" event, then select one or multiple input files for upload. Relative file paths are resolved relative to the current working directory. An empty list clears the selected files.
+    Examples:
+        upload_file('572', '/home/user/my_receipt.pdf')
+
+        upload_file('63', ['/home/bob/Documents/image.jpg', '/home/bob/Documents/file.zip'])
+"""
+
+
+for _, action in _browser_action_space.action_set.items():
+    assert (
+        action.signature in _BROWSER_TOOL_DESCRIPTION
+    ), f'Browser description mismatch. Please double check if the BrowserGym updated their action space.\n\nAction: {action.signature}'
+    assert (
+        action.description in _BROWSER_TOOL_DESCRIPTION
+    ), f'Browser description mismatch. Please double check if the BrowserGym updated their action space.\n\nAction: {action.description}'
+
+BrowserTool = ChatCompletionToolParam(
+    type='function',
+    function=ChatCompletionToolParamFunctionChunk(
+        name='browser',
+        description=_BROWSER_DESCRIPTION,
+        parameters={
+            'type': 'object',
+            'properties': {
+                'code': {
+                    'type': 'string',
+                    'description': (
+                        'The Python code that interacts with the browser.\n'
+                        + _BROWSER_TOOL_DESCRIPTION
+                    ),
+                }
+            },
+            'required': ['code'],
+        },
+    ),
+)
+
+_FINISH_DESCRIPTION = """Finish the interaction when the task is complete OR if the assistant cannot proceed further with the task."""
+
+FinishTool = ChatCompletionToolParam(
+    type='function',
+    function=ChatCompletionToolParamFunctionChunk(
+        name='finish',
+        description=_FINISH_DESCRIPTION,
+    ),
+)
+
+
+def combine_thought(action: Action, thought: str) -> Action:
+    if not hasattr(action, 'thought'):
+        return action
+    if thought:
+        action.thought = thought
+    return action
+
+
+def response_to_actions(response: ModelResponse) -> list[Action]:
+    actions: list[Action] = []
+    assert len(response.choices) == 1, 'Only one choice is supported for now'
+    assistant_msg = response.choices[0].message
+    if assistant_msg.tool_calls:
+        # Check if there's assistant_msg.content. If so, add it to the thought
+        thought = ''
+        if isinstance(assistant_msg.content, str):
+            thought = assistant_msg.content
+        elif isinstance(assistant_msg.content, list):
+            for msg in assistant_msg.content:
+                if msg['type'] == 'text':
+                    thought += msg['text']
+
+        # Process each tool call to OpenHands action
+        for i, tool_call in enumerate(assistant_msg.tool_calls):
+            action: Action
+            try:
+                arguments = json.loads(tool_call.function.arguments)
+            except json.decoder.JSONDecodeError as e:
+                raise RuntimeError(
+                    f'Failed to parse tool call arguments: {tool_call.function.arguments}'
+                ) from e
+            if tool_call.function.name == 'execute_bash':
+                action = CmdRunAction(**arguments)
+            elif tool_call.function.name == 'execute_ipython_cell':
+                action = IPythonRunCellAction(**arguments)
+            elif tool_call.function.name == 'delegate_to_browsing_agent':
+                action = AgentDelegateAction(
+                    agent='BrowsingAgent',
+                    inputs=arguments,
+                )
+            elif tool_call.function.name == 'finish':
+                action = AgentFinishAction()
+            elif tool_call.function.name == 'edit_file':
+                action = FileEditAction(**arguments)
+            elif tool_call.function.name == 'str_replace_editor':
+                # We implement this in agent_skills, which can be used via Jupyter
+                # convert tool_call.function.arguments to kwargs that can be passed to file_editor
+                code = f'print(file_editor(**{arguments}))'
+                logger.debug(
+                    f'TOOL CALL: str_replace_editor -> file_editor with code: {code}'
+                )
+                action = IPythonRunCellAction(code=code, include_extra=False)
+            elif tool_call.function.name == 'browser':
+                action = BrowseInteractiveAction(browser_actions=arguments['code'])
+            elif tool_call.function.name == 'web_read':
+                action = BrowseURLAction(url=arguments['url'])
+            else:
+                raise FunctionCallNotExistsError(
+                    f'Tool {tool_call.function.name} is not registered. (arguments: {arguments}). Please check the tool name and retry with an existing tool.'
+                )
+
+            # We only add thought to the first action
+            if i == 0:
+                action = combine_thought(action, thought)
+            # Add metadata for tool calling
+            action.tool_call_metadata = ToolCallMetadata(
+                tool_call_id=tool_call.id,
+                function_name=tool_call.function.name,
+                model_response=response,
+                total_calls_in_response=len(assistant_msg.tool_calls),
+            )
+            actions.append(action)
+    else:
+        actions.append(
+            MessageAction(content=assistant_msg.content, wait_for_response=True)
+        )
+
+    assert len(actions) >= 1
+    return actions
+
+
+def get_tools(
+    codeact_enable_browsing: bool = False,
+    codeact_enable_llm_editor: bool = False,
+    codeact_enable_jupyter: bool = False,
+) -> list[ChatCompletionToolParam]:
+    tools = [CmdRunTool, FinishTool]
+    if codeact_enable_browsing:
+        tools.append(WebReadTool)
+        tools.append(BrowserTool)
+    if codeact_enable_jupyter:
+        tools.append(IPythonTool)
+    if codeact_enable_llm_editor:
+        tools.append(LLMBasedFileEditTool)
+    else:
+        tools.append(StrReplaceEditorTool)
+    return tools

agenthub/codeact_agent/prompts/system_prompt.j2

@@ -0,0 +1,6 @@
+You are OpenHands agent, a helpful AI assistant that can interact with a computer to solve tasks.
+<IMPORTANT>
+* If user provides a path, you should NOT assume it's relative to the current working directory. Instead, you should explore the file system to find the file before working on it.
+* When configuring git credentials, use "openhands" as the user.name and "openhands@all-hands.dev" as the user.email by default, unless explicitly instructed otherwise.
+* The assistant MUST NOT include comments in the code unless they are necessary to describe non-obvious behavior.
+</IMPORTANT>

agenthub/dummy_agent/agent.py

@@ -1,4 +1,4 @@
-from typing import TypedDict
+from typing import TypedDict, Union
 
 from openhands.controller.agent import Agent
 from openhands.controller.state.state import State
@@ -6,6 +6,7 @@ from openhands.core.config import AgentConfig
 from openhands.core.schema import AgentState
 from openhands.events.action import (
     Action,
+    AddTaskAction,
     AgentFinishAction,
     AgentRejectAction,
     BrowseInteractiveAction,
@@ -14,14 +15,14 @@ from openhands.events.action import (
     FileReadAction,
     FileWriteAction,
     MessageAction,
+    ModifyTaskAction,
 )
 from openhands.events.observation import (
     AgentStateChangedObservation,
-    BrowserOutputObservation,
-    CmdOutputMetadata,
     CmdOutputObservation,
     FileReadObservation,
     FileWriteObservation,
+    NullObservation,
     Observation,
 )
 from openhands.events.serialization.event import event_to_dict
@@ -48,13 +49,31 @@ class DummyAgent(Agent):
     def __init__(self, llm: LLM, config: AgentConfig):
         super().__init__(llm, config)
         self.steps: list[ActionObs] = [
+            {
+                'action': AddTaskAction(
+                    parent='None', goal='check the current directory'
+                ),
+                'observations': [],
+            },
+            {
+                'action': AddTaskAction(parent='0', goal='run ls'),
+                'observations': [],
+            },
+            {
+                'action': ModifyTaskAction(task_id='0', state='in_progress'),
+                'observations': [],
+            },
             {
                 'action': MessageAction('Time to get started!'),
                 'observations': [],
             },
             {
                 'action': CmdRunAction(command='echo "foo"'),
-                'observations': [CmdOutputObservation('foo', command='echo "foo"')],
+                'observations': [
+                    CmdOutputObservation(
+                        'foo', command_id=-1, command='echo "foo"', exit_code=0
+                    )
+                ],
             },
             {
                 'action': FileWriteAction(
@@ -77,20 +96,16 @@ class DummyAgent(Agent):
                 'observations': [
                     CmdOutputObservation(
                         'bash: hello.sh: No such file or directory',
+                        command_id=-1,
                         command='bash workspace/hello.sh',
-                        metadata=CmdOutputMetadata(exit_code=127),
+                        exit_code=127,
                     )
                 ],
             },
             {
                 'action': BrowseURLAction(url='https://google.com'),
                 'observations': [
-                    BrowserOutputObservation(
-                        '<html><body>Simulated Google page</body></html>',
-                        url='https://google.com',
-                        screenshot='',
-                        trigger_by_action='',
-                    ),
+                    # BrowserOutputObservation('<html><body>Simulated Google page</body></html>',url='https://google.com',screenshot=''),
                 ],
             },
             {
@@ -98,17 +113,12 @@ class DummyAgent(Agent):
                     browser_actions='goto("https://google.com")'
                 ),
                 'observations': [
-                    BrowserOutputObservation(
-                        '<html><body>Simulated Google page after interaction</body></html>',
-                        url='https://google.com',
-                        screenshot='',
-                        trigger_by_action='',
-                    ),
+                    # BrowserOutputObservation('<html><body>Simulated Google page after interaction</body></html>',url='https://google.com',screenshot=''),
                 ],
             },
             {
                 'action': AgentRejectAction(),
-                'observations': [AgentStateChangedObservation('', AgentState.REJECTED)],
+                'observations': [NullObservation('')],
             },
             {
                 'action': AgentFinishAction(
@@ -125,6 +135,30 @@ class DummyAgent(Agent):
         current_step = self.steps[state.iteration]
         action = current_step['action']
 
+        # If the action is AddTaskAction or ModifyTaskAction, update the parent ID or task_id
+        if isinstance(action, AddTaskAction):
+            if action.parent == 'None':
+                action.parent = ''  # Root task has no parent
+            elif action.parent == '0':
+                action.parent = state.root_task.id
+            elif action.parent.startswith('0.'):
+                action.parent = f'{state.root_task.id}{action.parent[1:]}'
+        elif isinstance(action, ModifyTaskAction):
+            if action.task_id == '0':
+                action.task_id = state.root_task.id
+            elif action.task_id.startswith('0.'):
+                action.task_id = f'{state.root_task.id}{action.task_id[1:]}'
+            # Ensure the task_id doesn't start with a dot
+            if action.task_id.startswith('.'):
+                action.task_id = action.task_id[1:]
+        elif isinstance(action, (BrowseURLAction, BrowseInteractiveAction)):
+            try:
+                return self.simulate_browser_action(action)
+            except (
+                Exception
+            ):  # This could be a specific exception for browser unavailability
+                return self.handle_browser_unavailable(action)
+
         if state.iteration > 0:
             prev_step = self.steps[state.iteration - 1]
 
@@ -147,6 +181,8 @@ class DummyAgent(Agent):
                         obs.pop('timestamp', None)
                         obs.pop('cause', None)
                         obs.pop('source', None)
+                        if 'extras' in obs:
+                            obs['extras'].pop('command_id', None)
 
                     if hist_obs != expected_obs:
                         print(
@@ -154,3 +190,22 @@ class DummyAgent(Agent):
                         )
 
         return action
+
+    def simulate_browser_action(
+        self, action: Union[BrowseURLAction, BrowseInteractiveAction]
+    ) -> Action:
+        # Instead of simulating, we'll reject the browser action
+        return self.handle_browser_unavailable(action)
+
+    def handle_browser_unavailable(
+        self, action: Union[BrowseURLAction, BrowseInteractiveAction]
+    ) -> Action:
+        # Create a message action to inform that browsing is not available
+        message = 'Browser actions are not available in the DummyAgent environment.'
+        if isinstance(action, BrowseURLAction):
+            message += f' Unable to browse URL: {action.url}'
+        elif isinstance(action, BrowseInteractiveAction):
+            message += (
+                f' Unable to perform interactive browsing: {action.browser_actions}'
+            )
+        return MessageAction(content=message)

agenthub/micro/agent.py

@@ -6,11 +6,11 @@ from openhands.controller.agent import Agent
 from openhands.controller.state.state import State
 from openhands.core.config import AgentConfig
 from openhands.core.message import ImageContent, Message, TextContent
+from openhands.core.utils import json
 from openhands.events.action import Action
 from openhands.events.event import Event
 from openhands.events.serialization.action import action_from_dict
 from openhands.events.serialization.event import event_to_memory
-from openhands.io import json
 from openhands.llm.llm import LLM
 
 
@@ -22,7 +22,7 @@ def parse_response(orig_response: str) -> Action:
     return action_from_dict(action_dict)
 
 
-def to_json(obj: object, **kwargs: dict) -> str:
+def to_json(obj, **kwargs):
     """Serialize an object to str format"""
     return json.dumps(obj, **kwargs)
 
@@ -32,9 +32,7 @@ class MicroAgent(Agent):
     prompt = ''
     agent_definition: dict = {}
 
-    def history_to_json(
-        self, history: list[Event], max_events: int = 20, **kwargs: dict
-    ) -> str:
+    def history_to_json(self, history: list[Event], max_events: int = 20, **kwargs):
         """
         Serialize and simplify history to str format
         """
@@ -52,17 +50,13 @@ class MicroAgent(Agent):
         # history is in reverse order, let's fix it
         processed_history.reverse()
 
-        # everything starts with a message
-        # the first message is already in the prompt as the task
-        # TODO: so we don't need to include it in the history
-
         return json.dumps(processed_history, **kwargs)
 
     def __init__(self, llm: LLM, config: AgentConfig):
         super().__init__(llm, config)
         if 'name' not in self.agent_definition:
             raise ValueError('Agent definition must contain a name')
-        self.prompt_template = Environment(loader=BaseLoader()).from_string(self.prompt)
+        self.prompt_template = Environment(loader=BaseLoader).from_string(self.prompt)
         self.delegates = all_microagents.copy()
         del self.delegates[self.agent_definition['name']]
 
@@ -76,7 +70,7 @@ class MicroAgent(Agent):
             delegates=self.delegates,
             latest_user_message=last_user_message,
         )
-        content: list[TextContent | ImageContent] = [TextContent(text=prompt)]
+        content = [TextContent(text=prompt)]
         if self.llm.vision_is_active() and last_image_urls:
             content.append(ImageContent(image_urls=last_image_urls))
         message = Message(role='user', content=content)

agenthub/planner_agent/__init__.py

@@ -0,0 +1,4 @@
+from openhands.agenthub.planner_agent.agent import PlannerAgent
+from openhands.controller.agent import Agent
+
+Agent.register('PlannerAgent', PlannerAgent)

agenthub/planner_agent/agent.py

@@ -0,0 +1,53 @@
+from openhands.agenthub.planner_agent.prompt import get_prompt_and_images
+from openhands.agenthub.planner_agent.response_parser import PlannerResponseParser
+from openhands.controller.agent import Agent
+from openhands.controller.state.state import State
+from openhands.core.config import AgentConfig
+from openhands.core.message import ImageContent, Message, TextContent
+from openhands.events.action import Action, AgentFinishAction
+from openhands.llm.llm import LLM
+
+
+class PlannerAgent(Agent):
+    VERSION = '1.0'
+    """
+    The planner agent utilizes a special prompting strategy to create long term plans for solving problems.
+    The agent is given its previous action-observation pairs, current task, and hint based on last action taken at every step.
+    """
+    response_parser = PlannerResponseParser()
+
+    def __init__(self, llm: LLM, config: AgentConfig):
+        """Initialize the Planner Agent with an LLM
+
+        Parameters:
+        - llm (LLM): The llm to be used by this agent
+        """
+        super().__init__(llm, config)
+
+    def step(self, state: State) -> Action:
+        """Checks to see if current step is completed, returns AgentFinishAction if True.
+        Otherwise, creates a plan prompt and sends to model for inference, returning the result as the next action.
+
+        Parameters:
+        - state (State): The current state given the previous actions and observations
+
+        Returns:
+        - AgentFinishAction: If the last state was 'completed', 'verified', or 'abandoned'
+        - Action: The next action to take based on llm response
+        """
+        if state.root_task.state in [
+            'completed',
+            'verified',
+            'abandoned',
+        ]:
+            return AgentFinishAction()
+
+        prompt, image_urls = get_prompt_and_images(
+            state, self.llm.config.max_message_chars
+        )
+        content = [TextContent(text=prompt)]
+        if self.llm.vision_is_active() and image_urls:
+            content.append(ImageContent(image_urls=image_urls))
+        message = Message(role='user', content=content)
+        resp = self.llm.completion(messages=self.llm.format_messages_for_llm(message))
+        return self.response_parser.parse(resp)

agenthub/planner_agent/prompt.py

@@ -0,0 +1,191 @@
+from openhands.controller.state.state import State
+from openhands.core.logger import openhands_logger as logger
+from openhands.core.schema import ActionType
+from openhands.core.utils import json
+from openhands.events.action import (
+    Action,
+    NullAction,
+)
+from openhands.events.serialization.action import action_from_dict
+from openhands.events.serialization.event import event_to_memory
+
+HISTORY_SIZE = 20
+
+prompt = """
+# Task
+You're a diligent software engineer AI. You can't see, draw, or interact with a
+browser, but you can read and write files, and you can run commands, and you can think.
+
+You've been given the following task:
+
+%(task)s
+
+## Plan
+As you complete this task, you're building a plan and keeping
+track of your progress. Here's a JSON representation of your plan:
+
+%(plan)s
+
+
+%(plan_status)s
+
+You're responsible for managing this plan and the status of tasks in
+it, by using the `add_task` and `modify_task` actions described below.
+
+If the History below contradicts the state of any of these tasks, you
+MUST modify the task using the `modify_task` action described below.
+
+Be sure NOT to duplicate any tasks. Do NOT use the `add_task` action for
+a task that's already represented. Every task must be represented only once.
+
+Tasks that are sequential MUST be siblings. They must be added in order
+to their parent task.
+
+If you mark a task as 'completed', 'verified', or 'abandoned',
+all non-abandoned subtasks will be marked the same way.
+So before closing a task this way, you MUST not only be sure that it has
+been completed successfully--you must ALSO be sure that all its subtasks
+are ready to be marked the same way.
+
+If, and only if, ALL tasks have already been marked verified,
+you MUST respond with the `finish` action.
+
+## History
+Here is a recent history of actions you've taken in service of this plan,
+as well as observations you've made. This only includes the MOST RECENT
+ten actions--more happened before that.
+
+%(history)s
+
+
+Your most recent action is at the bottom of that history.
+
+## Action
+What is your next thought or action? Your response must be in JSON format.
+
+It must be an object, and it must contain two fields:
+* `action`, which is one of the actions below
+* `args`, which is a map of key-value pairs, specifying the arguments for that action
+
+* `read` - reads the content of a file. Arguments:
+  * `path` - the path of the file to read
+* `write` - writes the content to a file. Arguments:
+  * `path` - the path of the file to write
+  * `content` - the content to write to the file
+* `run` - runs a command on the command line in a Linux shell. Arguments:
+  * `command` - the command to run
+* `browse` - opens a web page. Arguments:
+  * `url` - the URL to open
+* `message` - make a plan, set a goal, record your thoughts, or ask for more input from the user. Arguments:
+  * `content` - the message to record
+  * `wait_for_response` - set to `true` to wait for the user to respond before proceeding
+* `add_task` - add a task to your plan. Arguments:
+  * `parent` - the ID of the parent task (leave empty if it should go at the top level)
+  * `goal` - the goal of the task
+  * `subtasks` - a list of subtasks, each of which is a map with a `goal` key.
+* `modify_task` - close a task. Arguments:
+  * `task_id` - the ID of the task to close
+  * `state` - set to 'in_progress' to start the task, 'completed' to finish it, 'verified' to assert that it was successful, 'abandoned' to give up on it permanently, or `open` to stop working on it for now.
+* `finish` - if ALL of your tasks and subtasks have been verified or abandoned, and you're absolutely certain that you've completed your task and have tested your work, use the finish action to stop working.
+
+You MUST take time to think in between read, write, run, and browse actions--do this with the `message` action.
+You should never act twice in a row without thinking. But if your last several
+actions are all `message` actions, you should consider taking a different action.
+
+What is your next thought or action? Again, you must reply with JSON, and only with JSON.
+
+%(hint)s
+"""
+
+
+def get_hint(latest_action_id: str) -> str:
+    """Returns action type hint based on given action_id"""
+    hints = {
+        '': "You haven't taken any actions yet. Start by using `ls` to check out what files you're working with.",
+        ActionType.RUN: 'You should think about the command you just ran, what output it gave, and how that affects your plan.',
+        ActionType.READ: 'You should think about the file you just read, what you learned from it, and how that affects your plan.',
+        ActionType.WRITE: 'You just changed a file. You should think about how it affects your plan.',
+        ActionType.BROWSE: 'You should think about the page you just visited, and what you learned from it.',
+        ActionType.MESSAGE: "Look at your last thought in the history above. What does it suggest? Don't think anymore--take action.",
+        ActionType.ADD_TASK: 'You should think about the next action to take.',
+        ActionType.MODIFY_TASK: 'You should think about the next action to take.',
+        ActionType.SUMMARIZE: '',
+        ActionType.FINISH: '',
+    }
+    return hints.get(latest_action_id, '')
+
+
+def get_prompt_and_images(
+    state: State, max_message_chars: int
+) -> tuple[str, list[str] | None]:
+    """Gets the prompt for the planner agent.
+
+    Formatted with the most recent action-observation pairs, current task, and hint based on last action
+
+    Parameters:
+    - state (State): The state of the current agent
+
+    Returns:
+    - str: The formatted string prompt with historical values
+    """
+    # the plan
+    plan_str = json.dumps(state.root_task.to_dict(), indent=2)
+
+    # the history
+    history_dicts = []
+    latest_action: Action = NullAction()
+
+    # retrieve the latest HISTORY_SIZE events
+    for event_count, event in enumerate(reversed(state.history)):
+        if event_count >= HISTORY_SIZE:
+            break
+        if latest_action == NullAction() and isinstance(event, Action):
+            latest_action = event
+        history_dicts.append(event_to_memory(event, max_message_chars))
+
+    # history_dicts is in reverse order, lets fix it
+    history_dicts.reverse()
+
+    # and get it as a JSON string
+    history_str = json.dumps(history_dicts, indent=2)
+
+    # the plan status
+    current_task = state.root_task.get_current_task()
+    if current_task is not None:
+        plan_status = f"You're currently working on this task:\n{current_task.goal}."
+        if len(current_task.subtasks) == 0:
+            plan_status += "\nIf it's not achievable AND verifiable with a SINGLE action, you MUST break it down into subtasks NOW."
+    else:
+        plan_status = "You're not currently working on any tasks. Your next action MUST be to mark a task as in_progress."
+
+    # the hint, based on the last action
+    hint = get_hint(event_to_memory(latest_action, max_message_chars).get('action', ''))
+    logger.debug('HINT:\n' + hint, extra={'msg_type': 'DETAIL'})
+
+    # the last relevant user message (the task)
+    message, image_urls = state.get_current_user_intent()
+
+    # finally, fill in the prompt
+    return prompt % {
+        'task': message,
+        'plan': plan_str,
+        'history': history_str,
+        'hint': hint,
+        'plan_status': plan_status,
+    }, image_urls
+
+
+def parse_response(response: str) -> Action:
+    """Parses the model output to find a valid action to take
+    Parameters:
+    - response (str): A response from the model that potentially contains an Action.
+
+    Returns:
+    - Action: A valid next action to perform from model output
+    """
+    action_dict = json.loads(response)
+    if 'contents' in action_dict:
+        # The LLM gets confused here. Might as well be robust
+        action_dict['content'] = action_dict.pop('contents')
+    action = action_from_dict(action_dict)
+    return action

agenthub/planner_agent/response_parser.py

@@ -0,0 +1,37 @@
+from openhands.controller.action_parser import ResponseParser
+from openhands.core.utils import json
+from openhands.events.action import (
+    Action,
+)
+from openhands.events.serialization.action import action_from_dict
+
+
+class PlannerResponseParser(ResponseParser):
+    def __init__(self):
+        super().__init__()
+
+    def parse(self, response: str) -> Action:
+        action_str = self.parse_response(response)
+        return self.parse_action(action_str)
+
+    def parse_response(self, response) -> str:
+        # get the next action from the response
+        return response['choices'][0]['message']['content']
+
+    def parse_action(self, action_str: str) -> Action:
+        """Parses a string to find an action within it
+
+        Parameters:
+        - response (str): The string to be parsed
+
+        Returns:
+        - Action: The action that was found in the response string
+        """
+        # attempt to load the JSON dict from the response
+        action_dict = json.loads(action_str)
+
+        if 'content' in action_dict:
+            # The LLM gets confused here. Might as well be robust
+            action_dict['contents'] = action_dict.pop('content')
+
+        return action_from_dict(action_dict)

build.sh

@@ -1,4 +0,0 @@
-#!/usr/bin/env bash
-set -e
-
-poetry build -v

config.template.toml

@@ -1,368 +0,0 @@
-###################### OpenHands Configuration Example ######################
-#
-# All settings have default values, so you only need to uncomment and
-# modify what you want to change
-# The fields within each section are sorted in alphabetical order.
-#
-##############################################################################
-
-#################################### Core ####################################
-# General core configurations
-##############################################################################
-[core]
-# API key for E2B
-#e2b_api_key = ""
-
-# API key for Modal
-#modal_api_token_id = ""
-#modal_api_token_secret = ""
-
-# API key for Daytona
-#daytona_api_key = ""
-
-# Daytona Target
-#daytona_target = ""
-
-# Base path for the workspace
-workspace_base = "./workspace"
-
-# Cache directory path
-#cache_dir = "/tmp/cache"
-
-# Reasoning effort for o1 models (low, medium, high, or not set)
-#reasoning_effort = "medium"
-
-# Debugging enabled
-#debug = false
-
-# Disable color in terminal output
-#disable_color = false
-
-# Enable saving and restoring the session when run from CLI
-#enable_cli_session = false
-
-# Path to store trajectories, can be a folder or a file
-# If it's a folder, the session id will be used as the file name
-#save_trajectory_path="./trajectories"
-
-# Path to replay a trajectory, must be a file path
-# If provided, trajectory will be loaded and replayed before the
-# agent responds to any user instruction
-#replay_trajectory_path = ""
-
-# File store path
-#file_store_path = "/tmp/file_store"
-
-# File store type
-#file_store = "memory"
-
-# List of allowed file extensions for uploads
-#file_uploads_allowed_extensions = [".*"]
-
-# Maximum file size for uploads, in megabytes
-#file_uploads_max_file_size_mb = 0
-
-# Maximum budget per task, 0.0 means no limit
-#max_budget_per_task = 0.0
-
-# Maximum number of iterations
-#max_iterations = 100
-
-# Path to mount the workspace in the sandbox
-#workspace_mount_path_in_sandbox = "/workspace"
-
-# Path to mount the workspace
-#workspace_mount_path = ""
-
-# Path to rewrite the workspace mount path to
-#workspace_mount_rewrite = ""
-
-# Run as openhands
-#run_as_openhands = true
-
-# Runtime environment
-#runtime = "docker"
-
-# Name of the default agent
-#default_agent = "CodeActAgent"
-
-# JWT secret for authentication
-#jwt_secret = ""
-
-# Restrict file types for file uploads
-#file_uploads_restrict_file_types = false
-
-# List of allowed file extensions for uploads
-#file_uploads_allowed_extensions = [".*"]
-
-# Whether to enable the default LLM summarizing condenser when no condenser is specified in config
-# When true, a LLMSummarizingCondenserConfig will be used as the default condenser
-# When false, a NoOpCondenserConfig (no summarization) will be used
-#enable_default_condenser = true
-
-#################################### LLM #####################################
-# Configuration for LLM models (group name starts with 'llm')
-# use 'llm' for the default LLM config
-##############################################################################
-[llm]
-# AWS access key ID
-#aws_access_key_id = ""
-
-# AWS region name
-#aws_region_name = ""
-
-# AWS secret access key
-#aws_secret_access_key = ""
-
-# API key to use (For Headless / CLI only -  In Web this is overridden by Session Init)
-api_key = ""
-
-# API base URL (For Headless / CLI only -  In Web this is overridden by Session Init)
-#base_url = ""
-
-# API version
-#api_version = ""
-
-# Cost per input token
-#input_cost_per_token = 0.0
-
-# Cost per output token
-#output_cost_per_token = 0.0
-
-# Custom LLM provider
-#custom_llm_provider = ""
-
-# Embedding API base URL
-#embedding_base_url = ""
-
-# Embedding deployment name
-#embedding_deployment_name = ""
-
-# Embedding model to use
-embedding_model = "local"
-
-# Maximum number of characters in an observation's content
-#max_message_chars = 10000
-
-# Maximum number of input tokens
-#max_input_tokens = 0
-
-# Maximum number of output tokens
-#max_output_tokens = 0
-
-# Model to use. (For Headless / CLI only -  In Web this is overridden by Session Init)
-model = "gpt-4o"
-
-# Number of retries to attempt when an operation fails with the LLM.
-# Increase this value to allow more attempts before giving up
-#num_retries = 8
-
-# Maximum wait time (in seconds) between retry attempts
-# This caps the exponential backoff to prevent excessively long
-#retry_max_wait = 120
-
-# Minimum wait time (in seconds) between retry attempts
-# This sets the initial delay before the first retry
-#retry_min_wait = 15
-
-# Multiplier for exponential backoff calculation
-# The wait time increases by this factor after each failed attempt
-# A value of 2.0 means each retry waits twice as long as the previous one
-#retry_multiplier = 2.0
-
-# Drop any unmapped (unsupported) params without causing an exception
-#drop_params = false
-
-# Modify params for litellm to do transformations like adding a default message, when a message is empty.
-# Note: this setting is global, unlike drop_params, it cannot be overridden in each call to litellm.
-#modify_params = true
-
-# Using the prompt caching feature if provided by the LLM and supported
-#caching_prompt = true
-
-# Base URL for the OLLAMA API
-#ollama_base_url = ""
-
-# Temperature for the API
-#temperature = 0.0
-
-# Timeout for the API
-#timeout = 0
-
-# Top p for the API
-#top_p = 1.0
-
-# If model is vision capable, this option allows to disable image processing (useful for cost reduction).
-#disable_vision = true
-
-# Custom tokenizer to use for token counting
-# https://docs.litellm.ai/docs/completion/token_usage
-#custom_tokenizer = ""
-
-# Whether to use native tool calling if supported by the model. Can be true, false, or None by default, which chooses the model's default behavior based on the evaluation.
-# ATTENTION: Based on evaluation, enabling native function calling may lead to worse results
-# in some scenarios. Use with caution and consider testing with your specific use case.
-# https://github.com/All-Hands-AI/OpenHands/pull/4711
-#native_tool_calling = None
-
-[llm.gpt4o-mini]
-api_key = ""
-model = "gpt-4o"
-
-
-#################################### Agent ###################################
-# Configuration for agents (group name starts with 'agent')
-# Use 'agent' for the default agent config
-# otherwise, group name must be `agent.<agent_name>` (case-sensitive), e.g.
-# agent.CodeActAgent
-##############################################################################
-[agent]
-
-# whether the browsing tool is enabled
-codeact_enable_browsing = true
-
-# whether the LLM draft editor is enabled
-codeact_enable_llm_editor = false
-
-# whether the IPython tool is enabled
-codeact_enable_jupyter = true
-
-# Name of the micro agent to use for this agent
-#micro_agent_name = ""
-
-# Memory enabled
-#memory_enabled = false
-
-# Memory maximum threads
-#memory_max_threads = 3
-
-# LLM config group to use
-#llm_config = 'your-llm-config-group'
-
-# Whether to use prompt extension (e.g., microagent, repo/runtime info) at all
-#enable_prompt_extensions = true
-
-# List of microagents to disable
-#disabled_microagents = []
-
-# Whether history should be truncated to continue the session when hitting LLM context
-# length limit
-enable_history_truncation = true
-
-[agent.RepoExplorerAgent]
-# Example: use a cheaper model for RepoExplorerAgent to reduce cost, especially
-# useful when an agent doesn't demand high quality but uses a lot of tokens
-llm_config = 'gpt3'
-
-#################################### Sandbox ###################################
-# Configuration for the sandbox
-##############################################################################
-[sandbox]
-# Sandbox timeout in seconds
-#timeout = 120
-
-# Sandbox user ID
-#user_id = 1000
-
-# Container image to use for the sandbox
-#base_container_image = "nikolaik/python-nodejs:python3.12-nodejs22"
-
-# Use host network
-#use_host_network = false
-
-# runtime extra build args
-#runtime_extra_build_args = ["--network=host", "--add-host=host.docker.internal:host-gateway"]
-
-# Enable auto linting after editing
-#enable_auto_lint = false
-
-# Whether to initialize plugins
-#initialize_plugins = true
-
-# Extra dependencies to install in the runtime image
-#runtime_extra_deps = ""
-
-# Environment variables to set at the launch of the runtime
-#runtime_startup_env_vars = {}
-
-# BrowserGym environment to use for evaluation
-#browsergym_eval_env = ""
-
-#################################### Security ###################################
-# Configuration for security features
-##############################################################################
-[security]
-
-# Enable confirmation mode (For Headless / CLI only -  In Web this is overridden by Session Init)
-#confirmation_mode = false
-
-# The security analyzer to use (For Headless / CLI only -  In Web this is overridden by Session Init)
-#security_analyzer = ""
-
-#################################### Condenser #################################
-# Condensers control how conversation history is managed and compressed when
-# the context grows too large. Each agent uses one condenser configuration.
-##############################################################################
-[condenser]
-# The type of condenser to use. Available options:
-# - "noop": No condensing, keeps full history (default)
-# - "observation_masking": Keeps full event structure but masks older observations
-# - "recent": Keeps only recent events and discards older ones
-# - "llm": Uses an LLM to summarize conversation history
-# - "amortized": Intelligently forgets older events while preserving important context
-# - "llm_attention": Uses an LLM to prioritize most relevant context
-type = "noop"
-
-# Examples for each condenser type (uncomment and modify as needed):
-
-# 1. NoOp Condenser - No additional settings needed
-#type = "noop"
-
-# 2. Observation Masking Condenser
-#type = "observation_masking"
-# Number of most-recent events where observations will not be masked
-#attention_window = 100
-
-# 3. Recent Events Condenser
-#type = "recent"
-# Number of initial events to always keep (typically includes task description)
-#keep_first = 1
-# Maximum number of events to keep in history
-#max_events = 100
-
-# 4. LLM Summarizing Condenser
-#type = "llm"
-# Reference to an LLM config to use for summarization
-#llm_config = "condenser"
-# Number of initial events to always keep (typically includes task description)
-#keep_first = 1
-# Maximum size of history before triggering summarization
-#max_size = 100
-
-# 5. Amortized Forgetting Condenser
-#type = "amortized"
-# Number of initial events to always keep (typically includes task description)
-#keep_first = 1
-# Maximum size of history before triggering forgetting
-#max_size = 100
-
-# 6. LLM Attention Condenser
-#type = "llm_attention"
-# Reference to an LLM config to use for attention scoring
-#llm_config = "condenser"
-# Number of initial events to always keep (typically includes task description)
-#keep_first = 1
-# Maximum size of history before triggering attention mechanism
-#max_size = 100
-
-# Example of a custom LLM configuration for condensers that require an LLM
-# If not provided, it falls back to the default LLM
-#[llm.condenser]
-#model = "gpt-4o"
-#temperature = 0.1
-#max_tokens = 1024
-
-#################################### Eval ####################################
-# Configuration for the evaluation, please refer to the specific evaluation
-# plugin for the available options
-##############################################################################

containers/README.md

@@ -1,12 +0,0 @@
-# Docker Containers
-
-Each folder here contains a Dockerfile, and a config.sh describing how to build
-the images and where to push them. These images are built and pushed in GitHub Actions
-by the `ghcr.yml` workflow.
-
-## Building Manually
-
-```bash
-docker build -f containers/app/Dockerfile -t openhands .
-docker build -f containers/sandbox/Dockerfile -t sandbox .
-```

containers/app/Dockerfile

@@ -1,98 +0,0 @@
-ARG OPENHANDS_BUILD_VERSION=dev
-FROM node:21.7.2-bookworm-slim AS frontend-builder
-
-WORKDIR /app
-
-COPY ./frontend/package.json frontend/package-lock.json ./
-RUN npm install -g npm@10.5.1
-RUN npm ci
-
-COPY ./frontend ./
-RUN npm run build
-
-FROM python:3.12.3-slim AS backend-builder
-
-WORKDIR /app
-ENV PYTHONPATH='/app'
-
-ENV POETRY_NO_INTERACTION=1 \
-    POETRY_VIRTUALENVS_IN_PROJECT=1 \
-    POETRY_VIRTUALENVS_CREATE=1 \
-    POETRY_CACHE_DIR=/tmp/poetry_cache
-
-RUN apt-get update -y \
-    && apt-get install -y curl make git build-essential \
-    && python3 -m pip install poetry==1.8.2  --break-system-packages
-
-COPY ./pyproject.toml ./poetry.lock ./
-RUN touch README.md
-RUN export POETRY_CACHE_DIR && poetry install --without evaluation,llama-index --no-root && rm -rf $POETRY_CACHE_DIR
-
-FROM python:3.12.3-slim AS openhands-app
-
-WORKDIR /app
-
-ARG OPENHANDS_BUILD_VERSION #re-declare for this section
-
-ENV RUN_AS_OPENHANDS=true
-# A random number--we need this to be different from the user's UID on the host machine
-ENV OPENHANDS_USER_ID=42420
-ENV SANDBOX_LOCAL_RUNTIME_URL=http://host.docker.internal
-ENV USE_HOST_NETWORK=false
-ENV WORKSPACE_BASE=/opt/workspace_base
-ENV OPENHANDS_BUILD_VERSION=$OPENHANDS_BUILD_VERSION
-ENV SANDBOX_USER_ID=0
-ENV FILE_STORE=local
-ENV FILE_STORE_PATH=/.openhands-state
-RUN mkdir -p $FILE_STORE_PATH
-RUN mkdir -p $WORKSPACE_BASE
-
-RUN apt-get update -y \
-    && apt-get install -y curl ssh sudo
-
-# Default is 1000, but OSX is often 501
-RUN sed -i 's/^UID_MIN.*/UID_MIN 499/' /etc/login.defs
-# Default is 60000, but we've seen up to 200000
-RUN sed -i 's/^UID_MAX.*/UID_MAX 1000000/' /etc/login.defs
-
-RUN groupadd app
-RUN useradd -l -m -u $OPENHANDS_USER_ID -s /bin/bash openhands && \
-    usermod -aG app openhands && \
-    usermod -aG sudo openhands && \
-    echo '%sudo ALL=(ALL) NOPASSWD:ALL' >> /etc/sudoers
-RUN chown -R openhands:app /app && chmod -R 770 /app
-RUN sudo chown -R openhands:app $WORKSPACE_BASE && sudo chmod -R 770 $WORKSPACE_BASE
-USER openhands
-
-ENV VIRTUAL_ENV=/app/.venv \
-    PATH="/app/.venv/bin:$PATH" \
-    PYTHONPATH='/app'
-
-COPY --chown=openhands:app --chmod=770 --from=backend-builder ${VIRTUAL_ENV} ${VIRTUAL_ENV}
-RUN playwright install --with-deps chromium
-
-COPY --chown=openhands:app --chmod=770 ./microagents ./microagents
-COPY --chown=openhands:app --chmod=770 ./openhands ./openhands
-COPY --chown=openhands:app --chmod=777 ./openhands/runtime/plugins ./openhands/runtime/plugins
-COPY --chown=openhands:app --chmod=770 ./openhands/agenthub ./openhands/agenthub
-COPY --chown=openhands:app ./pyproject.toml ./pyproject.toml
-COPY --chown=openhands:app ./poetry.lock ./poetry.lock
-COPY --chown=openhands:app ./README.md ./README.md
-COPY --chown=openhands:app ./MANIFEST.in ./MANIFEST.in
-COPY --chown=openhands:app ./LICENSE ./LICENSE
-
-# This is run as "openhands" user, and will create __pycache__ with openhands:openhands ownership
-RUN python openhands/core/download.py # No-op to download assets
-# Add this line to set group ownership of all files/directories not already in "app" group
-# openhands:openhands -> openhands:app
-RUN find /app \! -group app -exec chgrp app {} +
-
-COPY --chown=openhands:app --chmod=770 --from=frontend-builder /app/build ./frontend/build
-COPY --chown=openhands:app --chmod=770 ./containers/app/entrypoint.sh /app/entrypoint.sh
-
-USER root
-
-WORKDIR /app
-
-ENTRYPOINT ["/app/entrypoint.sh"]
-CMD ["uvicorn", "openhands.server.listen:app", "--host", "0.0.0.0", "--port", "3000"]

containers/app/config.sh

@@ -1,4 +0,0 @@
-DOCKER_REGISTRY=ghcr.io
-DOCKER_ORG=all-hands-ai
-DOCKER_IMAGE=openhands
-DOCKER_BASE_DIR="."

containers/app/entrypoint.sh

@@ -1,69 +0,0 @@
-#!/bin/bash
-set -eo pipefail
-
-echo "Starting OpenHands..."
-if [[ $NO_SETUP == "true" ]]; then
-  echo "Skipping setup, running as $(whoami)"
-  "$@"
-  exit 0
-fi
-
-if [ "$(id -u)" -ne 0 ]; then
-  echo "The OpenHands entrypoint.sh must run as root"
-  exit 1
-fi
-
-if [ -z "$SANDBOX_USER_ID" ]; then
-  echo "SANDBOX_USER_ID is not set"
-  exit 1
-fi
-
-if [ -z "$WORKSPACE_MOUNT_PATH" ]; then
-  # This is set to /opt/workspace in the Dockerfile. But if the user isn't mounting, we want to unset it so that OpenHands doesn't mount at all
-  unset WORKSPACE_BASE
-fi
-
-if [[ "$SANDBOX_USER_ID" -eq 0 ]]; then
-  echo "Running OpenHands as root"
-  export RUN_AS_OPENHANDS=false
-  mkdir -p /root/.cache/ms-playwright/
-  if [ -d "/home/openhands/.cache/ms-playwright/" ]; then
-    mv /home/openhands/.cache/ms-playwright/ /root/.cache/
-  fi
-  "$@"
-else
-  echo "Setting up enduser with id $SANDBOX_USER_ID"
-  if id "enduser" &>/dev/null; then
-    echo "User enduser already exists. Skipping creation."
-  else
-    if ! useradd -l -m -u $SANDBOX_USER_ID -s /bin/bash enduser; then
-      echo "Failed to create user enduser with id $SANDBOX_USER_ID. Moving openhands user."
-      incremented_id=$(($SANDBOX_USER_ID + 1))
-      usermod -u $incremented_id openhands
-      if ! useradd -l -m -u $SANDBOX_USER_ID -s /bin/bash enduser; then
-        echo "Failed to create user enduser with id $SANDBOX_USER_ID for a second time. Exiting."
-        exit 1
-      fi
-    fi
-  fi
-  usermod -aG app enduser
-  # get the user group of /var/run/docker.sock and set openhands to that group
-  DOCKER_SOCKET_GID=$(stat -c '%g' /var/run/docker.sock)
-  echo "Docker socket group id: $DOCKER_SOCKET_GID"
-  if getent group $DOCKER_SOCKET_GID; then
-    echo "Group with id $DOCKER_SOCKET_GID already exists"
-  else
-    echo "Creating group with id $DOCKER_SOCKET_GID"
-    groupadd -g $DOCKER_SOCKET_GID docker
-  fi
-
-  mkdir -p /home/enduser/.cache/huggingface/hub/
-  mkdir -p /home/enduser/.cache/ms-playwright/
-  if [ -d "/home/openhands/.cache/ms-playwright/" ]; then
-    mv /home/openhands/.cache/ms-playwright/ /home/enduser/.cache/
-  fi
-
-  usermod -aG $DOCKER_SOCKET_GID enduser
-  echo "Running as enduser"
-  su enduser /bin/bash -c "${*@Q}" # This magically runs any arguments passed to the script as a command
-fi

containers/build.sh

@@ -1,156 +0,0 @@
-#!/usr/bin/env bash
-set -eo pipefail
-
-# Initialize variables with default values
-image_name=""
-org_name=""
-push=0
-load=0
-tag_suffix=""
-
-# Function to display usage information
-usage() {
-    echo "Usage: $0 -i <image_name> [-o <org_name>] [--push] [--load] [-t <tag_suffix>]"
-    echo "  -i: Image name (required)"
-    echo "  -o: Organization name"
-    echo "  --push: Push the image"
-    echo "  --load: Load the image"
-    echo "  -t: Tag suffix"
-    exit 1
-}
-
-# Parse command-line options
-while [[ $# -gt 0 ]]; do
-    case $1 in
-        -i) image_name="$2"; shift 2 ;;
-        -o) org_name="$2"; shift 2 ;;
-        --push) push=1; shift ;;
-        --load) load=1; shift ;;
-        -t) tag_suffix="$2"; shift 2 ;;
-        *) usage ;;
-    esac
-done
-# Check if required arguments are provided
-if [[ -z "$image_name" ]]; then
-    echo "Error: Image name is required."
-    usage
-fi
-
-echo "Building: $image_name"
-tags=()
-
-OPENHANDS_BUILD_VERSION="dev"
-
-cache_tag_base="buildcache"
-cache_tag="$cache_tag_base"
-
-if [[ -n $RELEVANT_SHA ]]; then
-  git_hash=$(git rev-parse --short "$RELEVANT_SHA")
-  tags+=("$git_hash")
-  tags+=("$RELEVANT_SHA")
-fi
-
-if [[ -n $GITHUB_REF_NAME ]]; then
-  # check if ref name is a version number
-  if [[ $GITHUB_REF_NAME =~ ^[0-9]+\.[0-9]+\.[0-9]+$ ]]; then
-    major_version=$(echo "$GITHUB_REF_NAME" | cut -d. -f1)
-    minor_version=$(echo "$GITHUB_REF_NAME" | cut -d. -f1,2)
-    tags+=("$major_version" "$minor_version")
-    tags+=("latest")
-  fi
-  sanitized_ref_name=$(echo "$GITHUB_REF_NAME" | sed 's/[^a-zA-Z0-9.-]\+/-/g')
-  OPENHANDS_BUILD_VERSION=$sanitized_ref_name
-  sanitized_ref_name=$(echo "$sanitized_ref_name" | tr '[:upper:]' '[:lower:]') # lower case is required in tagging
-  tags+=("$sanitized_ref_name")
-  cache_tag+="-${sanitized_ref_name}"
-fi
-
-if [[ -n $tag_suffix ]]; then
-  cache_tag+="-${tag_suffix}"
-  for i in "${!tags[@]}"; do
-    tags[$i]="${tags[$i]}-$tag_suffix"
-  done
-fi
-
-echo "Tags: ${tags[@]}"
-
-if [[ "$image_name" == "openhands" ]]; then
-  dir="./containers/app"
-elif [[ "$image_name" == "runtime" ]]; then
-  dir="./containers/runtime"
-else
-  dir="./containers/$image_name"
-fi
-
-if [[ (! -f "$dir/Dockerfile") && "$image_name" != "runtime" ]]; then
-  # Allow runtime to be built without a Dockerfile
-  echo "No Dockerfile found"
-  exit 1
-fi
-if [[ ! -f "$dir/config.sh" ]]; then
-  echo "No config.sh found for Dockerfile"
-  exit 1
-fi
-
-source "$dir/config.sh"
-
-if [[ -n "$org_name" ]]; then
-  DOCKER_ORG="$org_name"
-fi
-
-# If $DOCKER_IMAGE_SOURCE_TAG is set, add it to the tags
-if [[ -n "$DOCKER_IMAGE_SOURCE_TAG" ]]; then
-  tags+=("$DOCKER_IMAGE_SOURCE_TAG")
-fi
-# If $DOCKER_IMAGE_TAG is set, add it to the tags
-if [[ -n "$DOCKER_IMAGE_TAG" ]]; then
-  tags+=("$DOCKER_IMAGE_TAG")
-fi
-
-DOCKER_REPOSITORY="$DOCKER_REGISTRY/$DOCKER_ORG/$DOCKER_IMAGE"
-DOCKER_REPOSITORY=${DOCKER_REPOSITORY,,} # lowercase
-echo "Repo: $DOCKER_REPOSITORY"
-echo "Base dir: $DOCKER_BASE_DIR"
-
-args=""
-for tag in "${tags[@]}"; do
-  args+=" -t $DOCKER_REPOSITORY:$tag"
-done
-
-if [[ $push -eq 1 ]]; then
-  args+=" --push"
-  args+=" --cache-to=type=registry,ref=$DOCKER_REPOSITORY:$cache_tag,mode=max"
-fi
-
-if [[ $load -eq 1 ]]; then
-  args+=" --load"
-fi
-
-echo "Args: $args"
-
-# Modify the platform selection based on --load flag
-if [[ $load -eq 1 ]]; then
-  # When loading, build only for the current platform
-  platform=$(docker version -f '{{.Server.Os}}/{{.Server.Arch}}')
-else
-  # For push or without load, build for multiple platforms
-  platform="linux/amd64,linux/arm64"
-fi
-
-echo "Building for platform(s): $platform"
-
-docker buildx build \
-  $args \
-  --build-arg OPENHANDS_BUILD_VERSION="$OPENHANDS_BUILD_VERSION" \
-  --cache-from=type=registry,ref=$DOCKER_REPOSITORY:$cache_tag \
-  --cache-from=type=registry,ref=$DOCKER_REPOSITORY:$cache_tag_base-main \
-  --platform $platform \
-  --provenance=false \
-  -f "$dir/Dockerfile" \
-  "$DOCKER_BASE_DIR"
-
-# If load was requested, print the loaded images
-if [[ $load -eq 1 ]]; then
-  echo "Local images built:"
-  docker images "$DOCKER_REPOSITORY" --format "{{.Repository}}:{{.Tag}}"
-fi

containers/dev/Dockerfile

@@ -1,124 +0,0 @@
-# syntax=docker/dockerfile:1
-
-###
-FROM ubuntu:22.04 AS dind
-
-# https://docs.docker.com/engine/install/ubuntu/
-RUN apt-get update && apt-get install -y \
-	ca-certificates \
-	curl \
-	&& install -m 0755 -d /etc/apt/keyrings \
-	&& curl -fsSL https://download.docker.com/linux/ubuntu/gpg -o /etc/apt/keyrings/docker.asc \
-	&& chmod a+r /etc/apt/keyrings/docker.asc \
-	&& echo \
-		"deb [arch=$(dpkg --print-architecture) signed-by=/etc/apt/keyrings/docker.asc] https://download.docker.com/linux/ubuntu \
-		$(. /etc/os-release && echo "$VERSION_CODENAME") stable" | tee /etc/apt/sources.list.d/docker.list > /dev/null
-
-RUN apt-get update && apt-get install -y \
-	docker-ce \
-	docker-ce-cli \
-	containerd.io \
-	docker-buildx-plugin \
-	docker-compose-plugin \
-	&& rm -rf /var/lib/apt/lists/* \
-	&& apt-get clean \
-	&& apt-get autoremove -y
-
-###
-FROM dind AS openhands
-
-ENV DEBIAN_FRONTEND=noninteractive
-
-#
-RUN apt-get update && apt-get install -y \
-	bash \
-    build-essential \
-    curl \
-	git \
-	git-lfs \
-	software-properties-common \
-	make \
-    netcat \
-    sudo \
-	wget \
-	&& rm -rf /var/lib/apt/lists/* \
-	&& apt-get clean \
-	&& apt-get autoremove -y
-
-# https://github.com/cli/cli/blob/trunk/docs/install_linux.md
-RUN curl -fsSL https://cli.github.com/packages/githubcli-archive-keyring.gpg | dd of=/usr/share/keyrings/githubcli-archive-keyring.gpg \
-	&& chmod go+r /usr/share/keyrings/githubcli-archive-keyring.gpg \
-	&& echo "deb [arch=$(dpkg --print-architecture) signed-by=/usr/share/keyrings/githubcli-archive-keyring.gpg] https://cli.github.com/packages stable main" | tee /etc/apt/sources.list.d/github-cli.list > /dev/null \
-	&& apt-get update && apt-get -y install \
-    gh \
-  && rm -rf /var/lib/apt/lists/* \
-  && apt-get clean \
-  && apt-get autoremove -y
-
-# Python 3.12
-RUN add-apt-repository ppa:deadsnakes/ppa \
-    && apt-get update \
-    && apt-get install -y python3.12 python3.12-venv python3.12-dev python3-pip \
-    && ln -s /usr/bin/python3.12 /usr/bin/python
-
-# NodeJS >= 18.17.1
-RUN curl -fsSL https://deb.nodesource.com/setup_18.x | bash - \
-    && apt-get install -y nodejs
-
-# Poetry >= 1.8
-RUN curl -fsSL https://install.python-poetry.org | python3.12 - \
-    && ln -s ~/.local/bin/poetry /usr/local/bin/poetry
-
-#
-RUN <<EOF
-#!/bin/bash
-printf "#!/bin/bash
-set +x
-uname -a
-docker --version
-gh --version | head -n 1
-git --version
-#
-python --version
-echo node `node --version`
-echo npm `npm --version`
-poetry --version
-netcat -h 2>&1 | head -n 1
-" > /version.sh
-chmod a+x /version.sh
-EOF
-
-###
-FROM openhands AS dev
-
-RUN apt-get update && apt-get install -y \
-	dnsutils \
-	file \
-	iproute2 \
-	jq \
-	lsof \
-	ripgrep \
-	silversearcher-ag \
-	vim \
-	&& rm -rf /var/lib/apt/lists/* \
-	&& apt-get clean \
-	&& apt-get autoremove -y
-
-WORKDIR /app
-
-# cache build dependencies
-RUN \
-  --mount=type=bind,source=./,target=/app/ \
-  <<EOF
-#!/bin/bash
-make -s clean
-make -s check-dependencies
-make -s install-python-dependencies
-
-# NOTE
-# node_modules are .dockerignore-d therefore not mountable
-# make -s install-frontend-dependencies
-EOF
-
-#
-CMD ["bash"]

containers/dev/README.md

@@ -1,57 +0,0 @@
-# Develop in Docker
-
-> [!WARNING]
-> This is not officially supported and may not work.
-
-Install [Docker](https://docs.docker.com/engine/install/) on your host machine and run:
-
-```bash
-make docker-dev
-# same as:
-cd ./containers/dev
-./dev.sh
-```
-
-It could take some time if you are running for the first time as Docker will pull all the  tools required for building OpenHands. The next time you run again, it should be instant.
-
-## Build and run
-
-If everything goes well, you should be inside a container after Docker finishes building the `openhands:dev` image similar to the following:
-
-```bash
-Build and run in Docker ...
-root@93fc0005fcd2:/app#
-```
-
-You may now proceed with the normal [build and run](../../Development.md) workflow as if you were on the host.
-
-## Make changes
-
-The source code on the host is mounted as `/app` inside docker. You may edit the files as usual either inside the Docker container or on your host with your favorite IDE/editors.
-
-The following are also mapped as readonly from your host:
-
-```yaml
-# host credentials
-- $HOME/.git-credentials:/root/.git-credentials:ro
-- $HOME/.gitconfig:/root/.gitconfig:ro
-- $HOME/.npmrc:/root/.npmrc:ro
-```
-
-## VSCode
-
-Alternatively, if you use VSCode, you could also [attach to the running container](https://code.visualstudio.com/docs/devcontainers/attach-container).
-
-See details for [developing in docker](https://code.visualstudio.com/docs/devcontainers/containers) or simply ask `OpenHands` ;-)
-
-## Rebuild dev image
-
-You could optionally pass additional options to the build script.
-
-```bash
-make docker-dev OPTIONS="--build"
-# or
-./containers/dev/dev.sh --build
-```
-
-See [docker compose run](https://docs.docker.com/reference/cli/docker/compose/run/) for more options.

containers/dev/compose.yml

@@ -1,38 +0,0 @@
-#
-services:
-  dev:
-    privileged: true
-    build:
-      context: ${OPENHANDS_WORKSPACE:-../../}
-      dockerfile: ./containers/dev/Dockerfile
-    image: openhands:dev
-    container_name: openhands-dev
-    environment:
-      - BACKEND_HOST=${BACKEND_HOST:-"0.0.0.0"}
-      - SANDBOX_API_HOSTNAME=host.docker.internal
-      #
-      - SANDBOX_RUNTIME_CONTAINER_IMAGE=${SANDBOX_RUNTIME_CONTAINER_IMAGE:-ghcr.io/all-hands-ai/runtime:0.27-nikolaik}
-      - SANDBOX_USER_ID=${SANDBOX_USER_ID:-1234}
-      - WORKSPACE_MOUNT_PATH=${WORKSPACE_BASE:-$PWD/workspace}
-    ports:
-      - "3000:3000"
-    extra_hosts:
-      - "host.docker.internal:host-gateway"
-    volumes:
-      - /var/run/docker.sock:/var/run/docker.sock
-      - ${WORKSPACE_BASE:-$PWD/workspace}:/opt/workspace_base
-      # source code
-      - ${OPENHANDS_WORKSPACE:-../../}:/app
-      # host credentials
-      - $HOME/.git-credentials:/root/.git-credentials:ro
-      - $HOME/.gitconfig:/root/.gitconfig:ro
-      - $HOME/.npmrc:/root/.npmrc:ro
-      # cache
-      - cache-data:/root/.cache
-    pull_policy: never
-    stdin_open: true
-    tty: true
-
-##
-volumes:
-  cache-data:

containers/dev/dev.sh

@@ -1,39 +0,0 @@
-#!/usr/bin/env bash
-set -o pipefail
-
-function get_docker() {
-    echo "Docker is required to build and run OpenHands."
-    echo "https://docs.docker.com/get-started/get-docker/"
-    exit 1
-}
-
-function check_tools() {
-	command -v docker &>/dev/null || get_docker
-}
-
-function exit_if_indocker() {
-    if [ -f /.dockerenv ]; then
-        echo "Running inside a Docker container. Exiting..."
-        exit 1
-    fi
-}
-
-#
-exit_if_indocker
-
-check_tools
-
-##
-OPENHANDS_WORKSPACE=$(git rev-parse --show-toplevel)
-
-cd "$OPENHANDS_WORKSPACE/containers/dev/" || exit 1
-
-##
-export BACKEND_HOST="0.0.0.0"
-#
-export SANDBOX_USER_ID=$(id -u)
-export WORKSPACE_BASE=${WORKSPACE_BASE:-$OPENHANDS_WORKSPACE/workspace}
-
-docker compose run --rm --service-ports "$@" dev
-
-##

containers/e2b-sandbox/Dockerfile

@@ -1,19 +0,0 @@
-FROM ubuntu:22.04
-
-# install basic packages
-RUN apt-get update && apt-get install -y \
-    curl \
-    wget \
-    git \
-    vim \
-    nano \
-    unzip \
-    zip \
-    python3 \
-    python3-pip \
-    python3-venv \
-    python3-dev \
-    build-essential \
-    openssh-server \
-    sudo \
-    && rm -rf /var/lib/apt/lists/*

containers/e2b-sandbox/README.md

@@ -1,15 +0,0 @@
-# How to build custom E2B sandbox for OpenHands
-
-[E2B](https://e2b.dev) is an [open-source](https://github.com/e2b-dev/e2b) secure cloud environment (sandbox) made for running AI-generated code and agents. E2B offers [Python](https://pypi.org/project/e2b/) and [JS/TS](https://www.npmjs.com/package/e2b) SDK to spawn and control these sandboxes.
-
-
-1. Install the CLI with NPM.
-    ```sh
-    npm install -g @e2b/cli@latest
-    ```
-    Full CLI API is [here](https://e2b.dev/docs/cli/installation).
-
-1. Build the sandbox
-  ```sh
-  e2b template build --dockerfile ./Dockerfile --name "openhands"
-  ```

containers/e2b-sandbox/e2b.toml

@@ -1,14 +0,0 @@
-# This is a config for E2B sandbox template.
-# You can use 'template_id' (785n69crgahmz0lkdw9h) or 'template_name (openhands) from this config to spawn a sandbox:
-
-# Python SDK
-# from e2b import Sandbox
-# sandbox = Sandbox(template='openhands')
-
-# JS SDK
-# import { Sandbox } from 'e2b'
-# const sandbox = await Sandbox.create({ template: 'openhands' })
-
-dockerfile = "Dockerfile"
-template_name = "openhands"
-template_id = "785n69crgahmz0lkdw9h"

containers/runtime/README.md

@@ -1,12 +0,0 @@
-# Dynamically constructed Dockerfile
-
-This folder builds a runtime image (sandbox), which will use a dynamically generated `Dockerfile`
-that depends on the `base_image` **AND** a [Python source distribution](https://docs.python.org/3.10/distutils/sourcedist.html) that is based on the current commit of `openhands`.
-
-The following command will generate a `Dockerfile` file for `nikolaik/python-nodejs:python3.12-nodejs22` (the default base image), an updated `config.sh` and the runtime source distribution files/folders into `containers/runtime`:
-
-```bash
-poetry run python3 openhands/runtime/utils/runtime_build.py \
-    --base_image nikolaik/python-nodejs:python3.12-nodejs22 \
-    --build_folder containers/runtime
-```

containers/runtime/config.sh

@@ -1,7 +0,0 @@
-DOCKER_REGISTRY=ghcr.io
-DOCKER_ORG=all-hands-ai
-DOCKER_BASE_DIR="./containers/runtime"
-DOCKER_IMAGE=runtime
-# These variables will be appended by the runtime_build.py script
-# DOCKER_IMAGE_TAG=
-# DOCKER_IMAGE_SOURCE_TAG=

controller/agent.py

@@ -11,9 +11,7 @@ from openhands.core.exceptions import (
 )
 from openhands.llm.llm import LLM
 from openhands.runtime.plugins import PluginRequirement
-
-if TYPE_CHECKING:
-    from openhands.utils.prompt import PromptManager
+from openhands.utils.prompt import PromptManager
 
 
 class Agent(ABC):
@@ -36,7 +34,7 @@ class Agent(ABC):
         self.llm = llm
         self.config = config
         self._complete = False
-        self.prompt_manager: 'PromptManager' | None = None
+        self.prompt_manager: PromptManager | None = None
 
     @property
     def complete(self) -> bool:

controller/agent_controller.py

@@ -5,23 +5,15 @@ import traceback
 from typing import Callable, ClassVar, Type
 
 import litellm
-from litellm.exceptions import (
-    BadRequestError,
-    ContextWindowExceededError,
-    OpenAIError,
-    RateLimitError,
-)
+from litellm.exceptions import BadRequestError, ContextWindowExceededError
 
 from openhands.controller.agent import Agent
-from openhands.controller.replay import ReplayManager
 from openhands.controller.state.state import State, TrafficControlState
 from openhands.controller.stuck import StuckDetector
 from openhands.core.config import AgentConfig, LLMConfig
 from openhands.core.exceptions import (
-    AgentStuckInLoopError,
     FunctionCallNotExistsError,
     FunctionCallValidationError,
-    LLMContextWindowExceedError,
     LLMMalformedActionError,
     LLMNoActionError,
     LLMResponseError,
@@ -33,6 +25,7 @@ from openhands.events import EventSource, EventStream, EventStreamSubscriber
 from openhands.events.action import (
     Action,
     ActionConfirmationStatus,
+    AddTaskAction,
     AgentDelegateAction,
     AgentFinishAction,
     AgentRejectAction,
@@ -40,19 +33,20 @@ from openhands.events.action import (
     CmdRunAction,
     IPythonRunCellAction,
     MessageAction,
+    ModifyTaskAction,
     NullAction,
 )
 from openhands.events.event import Event
 from openhands.events.observation import (
-    AgentCondensationObservation,
     AgentDelegateObservation,
     AgentStateChangedObservation,
     ErrorObservation,
     NullObservation,
     Observation,
 )
-from openhands.events.serialization.event import event_to_trajectory, truncate_content
+from openhands.events.serialization.event import truncate_content
 from openhands.llm.llm import LLM
+from openhands.utils.shutdown_listener import should_continue
 
 # note: RESUME is only available on web GUI
 TRAFFIC_CONTROL_REMINDER = (
@@ -69,6 +63,7 @@ class AgentController:
     confirmation_mode: bool
     agent_to_llm_config: dict[str, LLMConfig]
     agent_configs: dict[str, AgentConfig]
+    agent_task: asyncio.Future | None = None
     parent: 'AgentController | None' = None
     delegate: 'AgentController | None' = None
     _pending_action: Action | None = None
@@ -78,7 +73,6 @@ class AgentController:
         NullObservation,
         ChangeAgentStateAction,
         AgentStateChangedObservation,
-        AgentCondensationObservation,
     )
 
     def __init__(
@@ -95,7 +89,6 @@ class AgentController:
         is_delegate: bool = False,
         headless_mode: bool = True,
         status_callback: Callable | None = None,
-        replay_events: list[Event] | None = None,
     ):
         """Initializes a new instance of the AgentController class.
 
@@ -114,21 +107,17 @@ class AgentController:
             is_delegate: Whether this controller is a delegate.
             headless_mode: Whether the agent is run in headless mode.
             status_callback: Optional callback function to handle status updates.
-            replay_events: A list of logs to replay.
         """
+        self._step_lock = asyncio.Lock()
         self.id = sid
         self.agent = agent
         self.headless_mode = headless_mode
-        self.is_delegate = is_delegate
 
-        # the event stream must be set before maybe subscribing to it
+        # subscribe to the event stream
         self.event_stream = event_stream
-
-        # subscribe to the event stream if this is not a delegate
-        if not self.is_delegate:
-            self.event_stream.subscribe(
-                EventStreamSubscriber.AGENT_CONTROLLER, self.on_event, self.id
-            )
+        self.event_stream.subscribe(
+            EventStreamSubscriber.AGENT_CONTROLLER, self.on_event, self.id
+        )
 
         # state from the previous session, state from a parent agent, or a fresh state
         self.set_initial_state(
@@ -146,16 +135,12 @@ class AgentController:
         self._stuck_detector = StuckDetector(self.state)
         self.status_callback = status_callback
 
-        # replay-related
-        self._replay_manager = ReplayManager(replay_events)
-
-    async def close(self, set_stop_state=True) -> None:
+    async def close(self) -> None:
         """Closes the agent controller, canceling any ongoing tasks and unsubscribing from the event stream.
 
         Note that it's fairly important that this closes properly, otherwise the state is incomplete.
         """
-        if set_stop_state:
-            await self.set_agent_state_to(AgentState.STOPPED)
+        await self.set_agent_state_to(AgentState.STOPPED)
 
         # we made history, now is the time to rewrite it!
         # the final state.history will be used by external scripts like evals, tests, etc.
@@ -180,11 +165,7 @@ class AgentController:
         )
 
         # unsubscribe from the event stream
-        # only the root parent controller subscribes to the event stream
-        if not self.is_delegate:
-            self.event_stream.unsubscribe(
-                EventStreamSubscriber.AGENT_CONTROLLER, self.id
-            )
+        self.event_stream.unsubscribe(EventStreamSubscriber.AGENT_CONTROLLER, self.id)
         self._closed = True
 
     def log(self, level: str, message: str, extra: dict | None = None) -> None:
@@ -210,122 +191,42 @@ class AgentController:
         self,
         e: Exception,
     ):
-        """React to an exception by setting the agent state to error and sending a status message."""
         await self.set_agent_state_to(AgentState.ERROR)
         if self.status_callback is not None:
             err_id = ''
             if isinstance(e, litellm.AuthenticationError):
                 err_id = 'STATUS$ERROR_LLM_AUTHENTICATION'
-            elif isinstance(
-                e,
-                (
-                    litellm.ServiceUnavailableError,
-                    litellm.APIConnectionError,
-                    litellm.APIError,
-                ),
-            ):
-                err_id = 'STATUS$ERROR_LLM_SERVICE_UNAVAILABLE'
-            elif isinstance(e, litellm.InternalServerError):
-                err_id = 'STATUS$ERROR_LLM_INTERNAL_SERVER_ERROR'
-            elif isinstance(e, litellm.BadRequestError) and 'ExceededBudget' in str(e):
-                err_id = 'STATUS$ERROR_LLM_OUT_OF_CREDITS'
-            elif isinstance(e, RateLimitError):
-                await self.set_agent_state_to(AgentState.RATE_LIMITED)
-                return
-            self.status_callback('error', err_id, type(e).__name__ + ': ' + str(e))
+            self.status_callback('error', err_id, str(e))
 
-    def step(self):
-        asyncio.create_task(self._step_with_exception_handling())
+    async def start_step_loop(self):
+        """The main loop for the agent's step-by-step execution."""
+        self.log('info', 'Starting step loop...')
+        while True:
+            if not self._is_awaiting_observation() and not should_continue():
+                break
+            if self._closed:
+                break
+            try:
+                await self._step()
+            except asyncio.CancelledError:
+                self.log('debug', 'AgentController task was cancelled')
+                break
+            except Exception as e:
+                traceback.print_exc()
+                self.log('error', f'Error while running the agent: {e}')
+                await self._react_to_exception(e)
 
-    async def _step_with_exception_handling(self):
-        try:
-            await self._step()
-        except Exception as e:
-            self.log(
-                'error',
-                f'Error while running the agent (session ID: {self.id}): {e}. '
-                f'Traceback: {traceback.format_exc()}',
-            )
-            reported = RuntimeError(
-                'There was an unexpected error while running the agent. Please '
-                'report this error to the developers by opening an issue at '
-                'https://github.com/All-Hands-AI/OpenHands. Your session ID is '
-                f' {self.id}. Error type: {e.__class__.__name__}'
-            )
-            if (
-                isinstance(e, litellm.AuthenticationError)
-                or isinstance(e, litellm.BadRequestError)
-                or isinstance(e, RateLimitError)
-                or isinstance(e, LLMContextWindowExceedError)
-            ):
-                reported = e
-            await self._react_to_exception(reported)
+            await asyncio.sleep(0.1)
 
-    def should_step(self, event: Event) -> bool:
-        """
-        Whether the agent should take a step based on an event. In general,
-        the agent should take a step if it receives a message from the user,
-        or observes something in the environment (after acting).
-        """
-        # it might be the delegate's day in the sun
-        if self.delegate is not None:
-            return False
-
-        if isinstance(event, Action):
-            if isinstance(event, MessageAction) and event.source == EventSource.USER:
-                return True
-            if (
-                isinstance(event, MessageAction)
-                and self.get_agent_state() != AgentState.AWAITING_USER_INPUT
-            ):
-                # TODO: this is fragile, but how else to check if eligible?
-                return True
-            if isinstance(event, AgentDelegateAction):
-                return True
-            return False
-        if isinstance(event, Observation):
-            if isinstance(event, NullObservation) or isinstance(
-                event, AgentStateChangedObservation
-            ):
-                return False
-            return True
-        return False
-
-    def on_event(self, event: Event) -> None:
+    async def on_event(self, event: Event) -> None:
         """Callback from the event stream. Notifies the controller of incoming events.
 
         Args:
             event (Event): The incoming event to process.
         """
-
-        # If we have a delegate that is not finished or errored, forward events to it
-        if self.delegate is not None:
-            delegate_state = self.delegate.get_agent_state()
-            if delegate_state not in (
-                AgentState.FINISHED,
-                AgentState.ERROR,
-                AgentState.REJECTED,
-            ):
-                # Forward the event to delegate and skip parent processing
-                asyncio.get_event_loop().run_until_complete(
-                    self.delegate._on_event(event)
-                )
-                return
-            else:
-                # delegate is done or errored, so end it
-                self.end_delegate()
-                return
-
-        # continue parent processing only if there's no active delegate
-        asyncio.get_event_loop().run_until_complete(self._on_event(event))
-
-    async def _on_event(self, event: Event) -> None:
         if hasattr(event, 'hidden') and event.hidden:
             return
 
-        # Give others a little chance
-        await asyncio.sleep(0.01)
-
         # if the event is not filtered out, add it to the history
         if not any(isinstance(event, filter_type) for filter_type in self.filter_out):
             self.state.history.append(event)
@@ -335,27 +236,24 @@ class AgentController:
         elif isinstance(event, Observation):
             await self._handle_observation(event)
 
-        if self.should_step(event):
-            self.step()
-
     async def _handle_action(self, action: Action) -> None:
-        """Handles an Action from the agent or delegate."""
+        """Handles actions from the event stream.
+
+        Args:
+            action (Action): The action to handle.
+        """
         if isinstance(action, ChangeAgentStateAction):
             await self.set_agent_state_to(action.agent_state)  # type: ignore
         elif isinstance(action, MessageAction):
             await self._handle_message_action(action)
         elif isinstance(action, AgentDelegateAction):
             await self.start_delegate(action)
-            assert self.delegate is not None
-            # Post a MessageAction with the task for the delegate
-            if 'task' in action.inputs:
-                self.event_stream.add_event(
-                    MessageAction(content='TASK: ' + action.inputs['task']),
-                    EventSource.USER,
-                )
-                await self.delegate.set_agent_state_to(AgentState.RUNNING)
-            return
-
+        elif isinstance(action, AddTaskAction):
+            self.state.root_task.add_subtask(
+                action.parent, action.goal, action.subtasks
+            )
+        elif isinstance(action, ModifyTaskAction):
+            self.state.root_task.set_subtask_state(action.task_id, action.state)
         elif isinstance(action, AgentFinishAction):
             self.state.outputs = action.outputs
             self.state.metrics.merge(self.state.local_metrics)
@@ -435,28 +333,7 @@ class AgentController:
 
     def _reset(self) -> None:
         """Resets the agent controller"""
-        # make sure there is an Observation with the tool call metadata to be recognized by the agent
-        # otherwise the pending action is found in history, but it's incomplete without an obs with tool result
-        if self._pending_action and hasattr(self._pending_action, 'tool_call_metadata'):
-            # find out if there already is an observation with the same tool call metadata
-            found_observation = False
-            for event in self.state.history:
-                if (
-                    isinstance(event, Observation)
-                    and event.tool_call_metadata
-                    == self._pending_action.tool_call_metadata
-                ):
-                    found_observation = True
-                    break
 
-            # make a new ErrorObservation with the tool call metadata
-            if not found_observation:
-                obs = ErrorObservation(content='The action has not been executed.')
-                obs.tool_call_metadata = self._pending_action.tool_call_metadata
-                obs._cause = self._pending_action.id  # type: ignore[attr-defined]
-                self.event_stream.add_event(obs, EventSource.AGENT)
-
-        # reset the pending action, this will be called when the agent is STOPPED or ERROR
         self._pending_action = None
         self.agent.reset()
 
@@ -475,9 +352,6 @@ class AgentController:
             return
 
         if new_state in (AgentState.STOPPED, AgentState.ERROR):
-            # sync existing metrics BEFORE resetting the agent
-            await self.update_state_after_step()
-            self.state.metrics.merge(self.state.local_metrics)
             self._reset()
         elif (
             new_state == AgentState.RUNNING
@@ -523,6 +397,10 @@ class AgentController:
             EventSource.ENVIRONMENT,
         )
 
+        if new_state == AgentState.INIT and self.state.resume_state:
+            await self.set_agent_state_to(self.state.resume_state)
+            self.state.resume_state = None
+
     def get_agent_state(self) -> AgentState:
         """Returns the current state of the agent.
 
@@ -550,7 +428,7 @@ class AgentController:
         agent_cls: Type[Agent] = Agent.get_cls(action.agent)
         agent_config = self.agent_configs.get(action.agent, self.agent.config)
         llm_config = self.agent_to_llm_config.get(action.agent, self.agent.llm.config)
-        llm = LLM(config=llm_config, retry_listener=self._notify_on_llm_retry)
+        llm = LLM(config=llm_config)
         delegate_agent = agent_cls(llm=llm, config=agent_config)
         state = State(
             inputs=action.inputs or {},
@@ -568,7 +446,7 @@ class AgentController:
             f'start delegate, creating agent {delegate_agent.name} using LLM {llm}',
         )
 
-        # Create the delegate with is_delegate=True so it does NOT subscribe directly
+        self.event_stream.unsubscribe(EventStreamSubscriber.AGENT_CONTROLLER, self.id)
         self.delegate = AgentController(
             sid=self.id + '-delegate',
             agent=delegate_agent,
@@ -581,64 +459,25 @@ class AgentController:
             is_delegate=True,
             headless_mode=self.headless_mode,
         )
-
-    def end_delegate(self) -> None:
-        """Ends the currently active delegate (e.g., if it is finished or errored)
-        so that this controller can resume normal operation.
-        """
-        if self.delegate is None:
-            return
-
-        delegate_state = self.delegate.get_agent_state()
-
-        # update iteration that is shared across agents
-        self.state.iteration = self.delegate.state.iteration
-
-        # close the delegate controller before adding new events
-        asyncio.get_event_loop().run_until_complete(self.delegate.close())
-
-        if delegate_state in (AgentState.FINISHED, AgentState.REJECTED):
-            # retrieve delegate result
-            delegate_outputs = (
-                self.delegate.state.outputs if self.delegate.state else {}
-            )
-
-            # prepare delegate result observation
-            # TODO: replace this with AI-generated summary (#2395)
-            formatted_output = ', '.join(
-                f'{key}: {value}' for key, value in delegate_outputs.items()
-            )
-            content = (
-                f'{self.delegate.agent.name} finishes task with {formatted_output}'
-            )
-
-            # emit the delegate result observation
-            obs = AgentDelegateObservation(outputs=delegate_outputs, content=content)
-            self.event_stream.add_event(obs, EventSource.AGENT)
-        else:
-            # delegate state is ERROR
-            # emit AgentDelegateObservation with error content
-            delegate_outputs = (
-                self.delegate.state.outputs if self.delegate.state else {}
-            )
-            content = (
-                f'{self.delegate.agent.name} encountered an error during execution.'
-            )
-
-            # emit the delegate result observation
-            obs = AgentDelegateObservation(outputs=delegate_outputs, content=content)
-            self.event_stream.add_event(obs, EventSource.AGENT)
-
-        # unset delegate so parent can resume normal handling
-        self.delegate = None
-        self.delegateAction = None
+        await self.delegate.set_agent_state_to(AgentState.RUNNING)
 
     async def _step(self) -> None:
         """Executes a single step of the parent or delegate agent. Detects stuck agents and limits on the number of iterations and the task budget."""
         if self.get_agent_state() != AgentState.RUNNING:
+            await asyncio.sleep(1)
             return
 
         if self._pending_action:
+            await asyncio.sleep(1)
+            return
+
+        if self.delegate is not None:
+            assert self.delegate != self
+            if self.delegate.get_agent_state() == AgentState.PAUSED:
+                # no need to check too often
+                await asyncio.sleep(1)
+            else:
+                await self._delegate_step()
             return
 
         self.log(
@@ -647,6 +486,7 @@ class AgentController:
             extra={'msg_type': 'STEP'},
         )
 
+        # check if agent hit the resources limit
         stop_step = False
         if self.state.iteration >= self.state.max_iterations:
             stop_step = await self._handle_traffic_control(
@@ -659,61 +499,50 @@ class AgentController:
                     'budget', current_cost, self.max_budget_per_task
                 )
         if stop_step:
-            logger.warning('Stopping agent due to traffic control')
             return
 
         if self._is_stuck():
-            await self._react_to_exception(
-                AgentStuckInLoopError('Agent got stuck in a loop')
-            )
+            await self._react_to_exception(RuntimeError('Agent got stuck in a loop'))
             return
 
         self.update_state_before_step()
         action: Action = NullAction()
+        try:
+            action = self.agent.step(self.state)
+            if action is None:
+                raise LLMNoActionError('No action was returned')
+        except (
+            LLMMalformedActionError,
+            LLMNoActionError,
+            LLMResponseError,
+            FunctionCallValidationError,
+            FunctionCallNotExistsError,
+        ) as e:
+            self.event_stream.add_event(
+                ErrorObservation(
+                    content=str(e),
+                ),
+                EventSource.AGENT,
+            )
+            return
+        except (ContextWindowExceededError, BadRequestError) as e:
+            # FIXME: this is a hack until a litellm fix is confirmed
+            # Check if this is a nested context window error
+            error_str = str(e).lower()
+            if (
+                'contextwindowexceedederror' in error_str
+                or 'prompt is too long' in error_str
+                or isinstance(e, ContextWindowExceededError)
+            ):
+                # When context window is exceeded, keep roughly half of agent interactions
+                self.state.history = self._apply_conversation_window(self.state.history)
 
-        if self._replay_manager.should_replay():
-            # in replay mode, we don't let the agent to proceed
-            # instead, we replay the action from the replay trajectory
-            action = self._replay_manager.step()
-        else:
-            try:
-                action = self.agent.step(self.state)
-                if action is None:
-                    raise LLMNoActionError('No action was returned')
-                action._source = EventSource.AGENT  # type: ignore [attr-defined]
-            except (
-                LLMMalformedActionError,
-                LLMNoActionError,
-                LLMResponseError,
-                FunctionCallValidationError,
-                FunctionCallNotExistsError,
-            ) as e:
-                self.event_stream.add_event(
-                    ErrorObservation(
-                        content=str(e),
-                    ),
-                    EventSource.AGENT,
-                )
+                # Save the ID of the first event in our truncated history for future reloading
+                if self.state.history:
+                    self.state.start_id = self.state.history[0].id
+                # Don't add error event - let the agent retry with reduced context
                 return
-            except (ContextWindowExceededError, BadRequestError, OpenAIError) as e:
-                # FIXME: this is a hack until a litellm fix is confirmed
-                # Check if this is a nested context window error
-                # We have to rely on string-matching because LiteLLM doesn't consistently
-                # wrap the failure in a ContextWindowExceededError
-                error_str = str(e).lower()
-                if (
-                    'contextwindowexceedederror' in error_str
-                    or 'prompt is too long' in error_str
-                    or 'input length and `max_tokens` exceed context limit' in error_str
-                    or isinstance(e, ContextWindowExceededError)
-                ):
-                    if self.agent.config.enable_history_truncation:
-                        self._handle_long_context_error()
-                        return
-                    else:
-                        raise LLMContextWindowExceedError()
-                else:
-                    raise e
+            raise
 
         if action.runnable:
             if self.state.confirmation_mode and (
@@ -731,19 +560,74 @@ class AgentController:
                 == ActionConfirmationStatus.AWAITING_CONFIRMATION
             ):
                 await self.set_agent_state_to(AgentState.AWAITING_USER_CONFIRMATION)
-            self.event_stream.add_event(action, action._source)  # type: ignore [attr-defined]
+            self.event_stream.add_event(action, EventSource.AGENT)
 
         await self.update_state_after_step()
 
         log_level = 'info' if LOG_ALL_EVENTS else 'debug'
         self.log(log_level, str(action), extra={'msg_type': 'ACTION'})
 
-    def _notify_on_llm_retry(self, retries: int, max: int) -> None:
-        if self.status_callback is not None:
-            msg_id = 'STATUS$LLM_RETRY'
-            self.status_callback(
-                'info', msg_id, f'Retrying LLM request, {retries} / {max}'
+    async def _delegate_step(self) -> None:
+        """Executes a single step of the delegate agent."""
+        await self.delegate._step()  # type: ignore[union-attr]
+        assert self.delegate is not None
+        delegate_state = self.delegate.get_agent_state()
+        self.log('debug', f'Delegate state: {delegate_state}')
+        if delegate_state == AgentState.ERROR:
+            # update iteration that shall be shared across agents
+            self.state.iteration = self.delegate.state.iteration
+
+            # emit AgentDelegateObservation to mark delegate termination due to error
+            delegate_outputs = (
+                self.delegate.state.outputs if self.delegate.state else {}
             )
+            content = (
+                f'{self.delegate.agent.name} encountered an error during execution.'
+            )
+            obs = AgentDelegateObservation(outputs=delegate_outputs, content=content)
+            self.event_stream.add_event(obs, EventSource.AGENT)
+
+            # close the delegate upon error
+            await self.delegate.close()
+
+            # resubscribe parent when delegate is finished
+            self.event_stream.subscribe(
+                EventStreamSubscriber.AGENT_CONTROLLER, self.on_event, self.id
+            )
+            self.delegate = None
+            self.delegateAction = None
+
+        elif delegate_state in (AgentState.FINISHED, AgentState.REJECTED):
+            self.log('debug', 'Delegate agent has finished execution')
+            # retrieve delegate result
+            outputs = self.delegate.state.outputs if self.delegate.state else {}
+
+            # update iteration that shall be shared across agents
+            self.state.iteration = self.delegate.state.iteration
+
+            # close delegate controller: we must close the delegate controller before adding new events
+            await self.delegate.close()
+
+            # resubscribe parent when delegate is finished
+            self.event_stream.subscribe(
+                EventStreamSubscriber.AGENT_CONTROLLER, self.on_event, self.id
+            )
+
+            # update delegate result observation
+            # TODO: replace this with AI-generated summary (#2395)
+            formatted_output = ', '.join(
+                f'{key}: {value}' for key, value in outputs.items()
+            )
+            content = (
+                f'{self.delegate.agent.name} finishes task with {formatted_output}'
+            )
+            obs = AgentDelegateObservation(outputs=outputs, content=content)
+
+            # clean up delegate status
+            self.delegate = None
+            self.delegateAction = None
+            self.event_stream.add_event(obs, EventSource.AGENT)
+        return
 
     async def _handle_traffic_control(
         self, limit_type: str, current_value: float, max_value: float
@@ -812,20 +696,12 @@ class AgentController:
         # - the previous session, in which case it has history
         # - from a parent agent, in which case it has no history
         # - None / a new state
-
-        # If state is None, we create a brand new state and still load the event stream so we can restore the history
         if state is None:
             self.state = State(
                 inputs={},
                 max_iterations=max_iterations,
                 confirmation_mode=confirmation_mode,
             )
-            self.state.start_id = 0
-
-            self.log(
-                'debug',
-                f'AgentController {self.id} - created new state. start_id: {self.state.start_id}',
-            )
         else:
             self.state = state
 
@@ -837,13 +713,7 @@ class AgentController:
                 f'AgentController {self.id} initializing history from event {self.state.start_id}',
             )
 
-        # Always load from the event stream to avoid losing history
-        self._init_history()
-
-    def get_trajectory(self) -> list[dict]:
-        # state history could be partially hidden/truncated before controller is closed
-        assert self._closed
-        return [event_to_trajectory(event) for event in self.state.history]
+            self._init_history()
 
     def _init_history(self) -> None:
         """Initializes the agent's history from the event stream.
@@ -970,22 +840,6 @@ class AgentController:
         # make sure history is in sync
         self.state.start_id = start_id
 
-    def _handle_long_context_error(self) -> None:
-        # When context window is exceeded, keep roughly half of agent interactions
-        self.state.history = self._apply_conversation_window(self.state.history)
-
-        # Save the ID of the first event in our truncated history for future reloading
-        if self.state.history:
-            self.state.start_id = self.state.history[0].id
-
-        # Add an error event to trigger another step by the agent
-        self.event_stream.add_event(
-            AgentCondensationObservation(
-                content='Trimming prompt to meet context window limitations'
-            ),
-            EventSource.AGENT,
-        )
-
     def _apply_conversation_window(self, events: list[Event]) -> list[Event]:
         """Cuts history roughly in half when context window is exceeded, preserving action-observation pairs
         and ensuring the first user message is always included.
@@ -1086,12 +940,10 @@ class AgentController:
 
     def __repr__(self):
         return (
-            f'AgentController(id={getattr(self, "id", "<uninitialized>")}, '
-            f'agent={getattr(self, "agent", "<uninitialized>")!r}, '
-            f'event_stream={getattr(self, "event_stream", "<uninitialized>")!r}, '
-            f'state={getattr(self, "state", "<uninitialized>")!r}, '
-            f'delegate={getattr(self, "delegate", "<uninitialized>")!r}, '
-            f'_pending_action={getattr(self, "_pending_action", "<uninitialized>")!r})'
+            f'AgentController(id={self.id}, agent={self.agent!r}, '
+            f'event_stream={self.event_stream!r}, '
+            f'state={self.state!r}, agent_task={self.agent_task!r}, '
+            f'delegate={self.delegate!r}, _pending_action={self._pending_action!r})'
         )
 
     def _is_awaiting_observation(self):

controller/stuck.py

@@ -5,7 +5,7 @@ from openhands.events.action.commands import IPythonRunCellAction
 from openhands.events.action.empty import NullAction
 from openhands.events.action.message import MessageAction
 from openhands.events.event import Event, EventSource
-from openhands.events.observation import (
+from openhands.events.observation.commands import (
     CmdOutputObservation,
     IPythonRunCellObservation,
 )

core/cli.py

@@ -1,26 +1,24 @@
 import asyncio
 import logging
 import sys
+from typing import Type
 from uuid import uuid4
 
 from termcolor import colored
 
 import openhands.agenthub  # noqa F401 (we import this to get the agents registered)
+from openhands import __version__
+from openhands.controller import AgentController
+from openhands.controller.agent import Agent
 from openhands.core.config import (
     AppConfig,
-    parse_arguments,
-    setup_config_from_args,
+    get_parser,
+    load_app_config,
 )
 from openhands.core.logger import openhands_logger as logger
 from openhands.core.loop import run_agent_until_done
 from openhands.core.schema import AgentState
-from openhands.core.setup import (
-    create_agent,
-    create_controller,
-    create_runtime,
-    initialize_repository_for_runtime,
-)
-from openhands.events import EventSource, EventStreamSubscriber
+from openhands.events import EventSource, EventStream, EventStreamSubscriber
 from openhands.events.action import (
     Action,
     ActionConfirmationStatus,
@@ -34,8 +32,13 @@ from openhands.events.observation import (
     AgentStateChangedObservation,
     CmdOutputObservation,
     FileEditObservation,
+    NullObservation,
 )
-from openhands.io import read_input, read_task
+from openhands.llm.llm import LLM
+from openhands.runtime import get_runtime_cls
+from openhands.runtime.base import Runtime
+from openhands.security import SecurityAnalyzer, options
+from openhands.storage import get_file_store
 
 
 def display_message(message: str):
@@ -88,42 +91,68 @@ def display_event(event: Event, config: AppConfig):
         display_confirmation(event.confirmation_state)
 
 
-async def main(loop: asyncio.AbstractEventLoop):
-    """Runs the agent in CLI mode."""
+async def main():
+    """Runs the agent in CLI mode"""
 
-    args = parse_arguments()
+    parser = get_parser()
+    # Add the version argument
+    parser.add_argument(
+        '-v',
+        '--version',
+        action='version',
+        version=f'{__version__}',
+        help='Show the version number and exit',
+        default=None,
+    )
+    args = parser.parse_args()
+
+    if args.version:
+        print(f'OpenHands version: {__version__}')
+        return
 
     logger.setLevel(logging.WARNING)
+    config = load_app_config(config_file=args.config_file)
+    sid = 'cli'
 
-    # Load config from toml and override with command line arguments
-    config: AppConfig = setup_config_from_args(args)
-
-    # Read task from file, CLI args, or stdin
-    task_str = read_task(args, config.cli_multiline_input)
-
-    # If we have a task, create initial user action
-    initial_user_action = MessageAction(content=task_str) if task_str else None
-
-    sid = str(uuid4())
-    display_message(f'Session ID: {sid}')
-
-    agent = create_agent(config)
-
-    runtime = create_runtime(
-        config,
-        sid=sid,
-        headless_mode=True,
-        agent=agent,
+    agent_cls: Type[Agent] = Agent.get_cls(config.default_agent)
+    agent_config = config.get_agent_config(config.default_agent)
+    llm_config = config.get_llm_config_from_agent(config.default_agent)
+    agent = agent_cls(
+        llm=LLM(config=llm_config),
+        config=agent_config,
     )
 
-    controller, _ = create_controller(agent, runtime, config)
+    file_store = get_file_store(config.file_store, config.file_store_path)
+    event_stream = EventStream(sid, file_store)
 
-    event_stream = runtime.event_stream
+    runtime_cls = get_runtime_cls(config.runtime)
+    runtime: Runtime = runtime_cls(  # noqa: F841
+        config=config,
+        event_stream=event_stream,
+        sid=sid,
+        plugins=agent_cls.sandbox_plugins,
+        headless_mode=True,
+    )
+
+    if config.security.security_analyzer:
+        options.SecurityAnalyzers.get(
+            config.security.security_analyzer, SecurityAnalyzer
+        )(event_stream)
+
+    controller = AgentController(
+        agent=agent,
+        max_iterations=config.max_iterations,
+        max_budget_per_task=config.max_budget_per_task,
+        agent_to_llm_config=config.get_agent_to_llm_config_map(),
+        event_stream=event_stream,
+        confirmation_mode=config.security.confirmation_mode,
+    )
 
     async def prompt_for_next_task():
         # Run input() in a thread pool to avoid blocking the event loop
+        loop = asyncio.get_event_loop()
         next_message = await loop.run_in_executor(
-            None, read_input, config.cli_multiline_input
+            None, lambda: input('How can I help? >> ')
         )
         if not next_message.strip():
             await prompt_for_next_task()
@@ -136,12 +165,13 @@ async def main(loop: asyncio.AbstractEventLoop):
         event_stream.add_event(action, EventSource.USER)
 
     async def prompt_for_user_confirmation():
+        loop = asyncio.get_event_loop()
         user_confirmation = await loop.run_in_executor(
             None, lambda: input('Confirm action (possible security risk)? (y/n) >> ')
         )
         return user_confirmation.lower() == 'y'
 
-    async def on_event_async(event: Event):
+    async def on_event(event: Event):
         display_event(event, config)
         if isinstance(event, AgentStateChangedObservation):
             if event.agent_state in [
@@ -149,40 +179,25 @@ async def main(loop: asyncio.AbstractEventLoop):
                 AgentState.FINISHED,
             ]:
                 await prompt_for_next_task()
-            if event.agent_state == AgentState.AWAITING_USER_CONFIRMATION:
-                user_confirmed = await prompt_for_user_confirmation()
-                if user_confirmed:
-                    event_stream.add_event(
-                        ChangeAgentStateAction(AgentState.USER_CONFIRMED),
-                        EventSource.USER,
-                    )
-                else:
-                    event_stream.add_event(
-                        ChangeAgentStateAction(AgentState.USER_REJECTED),
-                        EventSource.USER,
-                    )
-
-    def on_event(event: Event) -> None:
-        loop.create_task(on_event_async(event))
+        if (
+            isinstance(event, NullObservation)
+            and controller.state.agent_state == AgentState.AWAITING_USER_CONFIRMATION
+        ):
+            user_confirmed = await prompt_for_user_confirmation()
+            if user_confirmed:
+                event_stream.add_event(
+                    ChangeAgentStateAction(AgentState.USER_CONFIRMED), EventSource.USER
+                )
+            else:
+                event_stream.add_event(
+                    ChangeAgentStateAction(AgentState.USER_REJECTED), EventSource.USER
+                )
 
     event_stream.subscribe(EventStreamSubscriber.MAIN, on_event, str(uuid4()))
 
     await runtime.connect()
 
-    # Initialize repository if needed
-    if config.sandbox.selected_repo:
-        initialize_repository_for_runtime(
-            runtime,
-            agent=agent,
-            selected_repository=config.sandbox.selected_repo,
-        )
-
-    if initial_user_action:
-        # If there's an initial user action, enqueue it and do not prompt again
-        event_stream.add_event(initial_user_action, EventSource.USER)
-    else:
-        # Otherwise prompt for the user's first message right away
-        asyncio.create_task(prompt_for_next_task())
+    asyncio.create_task(prompt_for_next_task())
 
     await run_agent_until_done(
         controller, runtime, [AgentState.STOPPED, AgentState.ERROR]
@@ -193,7 +208,7 @@ if __name__ == '__main__':
     loop = asyncio.new_event_loop()
     asyncio.set_event_loop(loop)
     try:
-        loop.run_until_complete(main(loop))
+        loop.run_until_complete(main())
     except KeyboardInterrupt:
         print('Received keyboard interrupt, shutting down...')
     except ConnectionRefusedError as e:

core/config/__init__.py

@@ -5,20 +5,17 @@ from openhands.core.config.config_utils import (
     OH_MAX_ITERATIONS,
     get_field_info,
 )
-from openhands.core.config.extended_config import ExtendedConfig
 from openhands.core.config.llm_config import LLMConfig
 from openhands.core.config.sandbox_config import SandboxConfig
 from openhands.core.config.security_config import SecurityConfig
 from openhands.core.config.utils import (
     finalize_config,
-    get_agent_config_arg,
     get_llm_config_arg,
     get_parser,
     load_app_config,
     load_from_env,
     load_from_toml,
     parse_arguments,
-    setup_config_from_args,
 )
 
 __all__ = [
@@ -29,15 +26,12 @@ __all__ = [
     'LLMConfig',
     'SandboxConfig',
     'SecurityConfig',
-    'ExtendedConfig',
     'load_app_config',
     'load_from_env',
     'load_from_toml',
     'finalize_config',
-    'get_agent_config_arg',
     'get_llm_config_arg',
     'get_field_info',
     'get_parser',
     'parse_arguments',
-    'setup_config_from_args',
 ]