Update pyproject.toml

Co-authored-by: openhands <openhands@all-hands.dev>
Co-authored-by: rohitvinodmalhotra@gmail.com <rohitvinodmalhotra@gmail.com>

.devcontainer/setup.sh

@@ -7,5 +7,8 @@ git config --global --add safe.directory "$(realpath .)"
 # Install `nc`
 sudo apt update && sudo apt install netcat -y
 
+# Install `uv` and `uvx`
+wget -qO- https://astral.sh/uv/install.sh | sh
+
 # Do common setup tasks
 source .openhands/setup.sh

.github/scripts/check_version_consistency.py

@@ -1,73 +0,0 @@
-#!/usr/bin/env python3
-import os
-import re
-import sys
-
-
-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 and docs/build directory
-        if '.git' in root or 'docs/build' 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)
-                        if matches:
-                            print(f'Found openhands version {matches} in {file_path}')
-                            openhands_versions.update(matches)
-
-                        # Find all runtime version references
-                        matches = version_pattern_runtime.findall(content)
-                        if matches:
-                            print(f'Found runtime version {matches} in {file_path}')
-                            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__), '..', '..'))
-    print(f'Checking version consistency in {repo_root}')
-    openhands_versions, runtime_versions = find_version_references(repo_root)
-
-    print(f'Found openhands versions: {sorted(openhands_versions)}')
-    print(f'Found runtime versions: {sorted(runtime_versions)}')
-
-    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/scripts/update_pr_description.sh

@@ -13,9 +13,9 @@ 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/openhands/runtime:${SHORT_SHA}-nikolaik \
+  -e SANDBOX_RUNTIME_CONTAINER_IMAGE=docker.openhands.dev/openhands/runtime:${SHORT_SHA}-nikolaik \
   --name openhands-app-${SHORT_SHA} \
-  docker.all-hands.dev/openhands/openhands:${SHORT_SHA}"
+  docker.openhands.dev/openhands/openhands:${SHORT_SHA}"
 
 # Define the uvx command
 UVX_RUN_COMMAND="uvx --python 3.12 --from git+https://github.com/OpenHands/OpenHands@${BRANCH_NAME}#subdirectory=openhands-cli openhands"

.github/workflows/cli-build-binary-and-optionally-release.yml

@@ -71,6 +71,14 @@ jobs:
 
           echo "✅ Build & test finished without ❌ markers"
 
+      - name: Verify binary files exist
+        run: |
+          if ! ls openhands-cli/dist/openhands* 1> /dev/null 2>&1; then
+            echo "❌ No binaries found to upload!"
+            exit 1
+          fi
+          echo "✅ Found binaries to upload."
+
       - name: Upload binary artifact
         uses: actions/upload-artifact@v4
         with:

.github/workflows/ghcr-build.yml

@@ -86,7 +86,7 @@ jobs:
 
   # Builds the runtime Docker images
   ghcr_build_runtime:
-    name: Build Image
+    name: Build Runtime Image
     runs-on: blacksmith-8vcpu-ubuntu-2204
     if: "!(github.event_name == 'push' && startsWith(github.ref, 'refs/tags/ext-v'))"
     permissions:
@@ -256,7 +256,7 @@ jobs:
   test_runtime_root:
     name: RT Unit Tests (Root)
     needs: [ghcr_build_runtime, define-matrix]
-    runs-on: blacksmith-8vcpu-ubuntu-2204
+    runs-on: blacksmith-4vcpu-ubuntu-2404
     strategy:
       fail-fast: false
       matrix:
@@ -298,7 +298,7 @@ jobs:
           # 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
+          # Install to be able to retry on failures for flakey tests
           poetry run pip install pytest-rerunfailures
 
           image_name=ghcr.io/${{ env.REPO_OWNER }}/runtime:${{ env.RELEVANT_SHA }}-${{ matrix.base_image.tag }}
@@ -311,14 +311,14 @@ jobs:
           SANDBOX_RUNTIME_CONTAINER_IMAGE=$image_name \
           TEST_IN_CI=true \
           RUN_AS_OPENHANDS=false \
-          poetry run pytest -n 0 -raRs --reruns 2 --reruns-delay 5 -s ./tests/runtime --ignore=tests/runtime/test_browsergym_envs.py --durations=10
+          poetry run pytest -n 5 -raRs --reruns 2 --reruns-delay 3 -s ./tests/runtime --ignore=tests/runtime/test_browsergym_envs.py --durations=10
         env:
           DEBUG: "1"
 
   # Run unit tests with the Docker runtime Docker images as openhands user
   test_runtime_oh:
     name: RT Unit Tests (openhands)
-    runs-on: blacksmith-8vcpu-ubuntu-2204
+    runs-on: blacksmith-4vcpu-ubuntu-2404
     needs: [ghcr_build_runtime, define-matrix]
     strategy:
       matrix:
@@ -370,7 +370,7 @@ jobs:
           SANDBOX_RUNTIME_CONTAINER_IMAGE=$image_name \
           TEST_IN_CI=true \
           RUN_AS_OPENHANDS=true \
-          poetry run pytest -n 0 -raRs --reruns 2 --reruns-delay 5 -s ./tests/runtime --ignore=tests/runtime/test_browsergym_envs.py --durations=10
+          poetry run pytest -n 5 -raRs --reruns 2 --reruns-delay 3 -s ./tests/runtime --ignore=tests/runtime/test_browsergym_envs.py --durations=10
         env:
           DEBUG: "1"
 

.github/workflows/lint.yml

@@ -90,16 +90,3 @@ jobs:
       - name: Run pre-commit hooks
         working-directory: ./openhands-cli
         run: pre-commit run --all-files --config ./dev_config/python/.pre-commit-config.yaml
-
-  # Check version consistency across documentation
-  check-version-consistency:
-    name: Check version consistency
-    runs-on: blacksmith-4vcpu-ubuntu-2204
-    steps:
-      - uses: actions/checkout@v4
-      - name: Set up python
-        uses: useblacksmith/setup-python@v6
-        with:
-          python-version: 3.12
-      - name: Run version consistency check
-        run: .github/scripts/check_version_consistency.py

.github/workflows/py-tests.yml

@@ -48,7 +48,10 @@ jobs:
           python-version: ${{ matrix.python-version }}
           cache: "poetry"
       - name: Install Python dependencies using Poetry
-        run: poetry install --with dev,test,runtime
+        run: |
+          poetry install --with dev,test,runtime
+          poetry run pip install pytest-xdist
+          poetry run pip install pytest-rerunfailures
       - name: Build Environment
         run: make build
       - name: Run Unit Tests
@@ -56,7 +59,7 @@ jobs:
         env:
           COVERAGE_FILE: ".coverage.${{ matrix.python_version }}"
       - name: Run Runtime Tests with CLIRuntime
-        run: PYTHONPATH=".:$PYTHONPATH" TEST_RUNTIME=cli poetry run pytest -s tests/runtime/test_bash.py --cov=openhands --cov-branch
+        run: PYTHONPATH=".:$PYTHONPATH" TEST_RUNTIME=cli poetry run pytest -n 5 --reruns 2 --reruns-delay 3 -s tests/runtime/test_bash.py --cov=openhands --cov-branch
         env:
           COVERAGE_FILE: ".coverage.runtime.${{ matrix.python_version }}"
       - name: Store coverage file
@@ -88,7 +91,7 @@ jobs:
       - name: Install Python dependencies using Poetry
         run: poetry install --with dev,test,runtime
       - name: Run Windows unit tests
-        run: poetry run pytest -svv tests/unit/runtime/utils/test_windows_bash.py
+        run: poetry run pytest -svv tests/runtime//test_windows_bash.py
         env:
           PYTHONPATH: ".;$env:PYTHONPATH"
           DEBUG: "1"
@@ -173,9 +176,9 @@ jobs:
           path: ".coverage.openhands-cli.${{ matrix.python-version }}"
           include-hidden-files: true
 
-
   coverage-comment:
     name: Coverage Comment
+    if: github.event_name == 'pull_request'
     runs-on: ubuntu-latest
     needs: [test-on-linux, test-enterprise, test-cli-python]
 
@@ -195,23 +198,8 @@ jobs:
         run: ln -sf openhands-cli/openhands_cli openhands_cli
 
       - name: Coverage comment
-        # In PR mode leaves a comment, otherwise records coverage in branch python-coverage-comment-action-data.
         id: coverage_comment
         uses: py-cov-action/python-coverage-comment-action@v3
         with:
           GITHUB_TOKEN: ${{ github.token }}
           MERGE_COVERAGE_FILES: true
-
-      - name: Enforce coverage
-        # Fail if on PR AND there are uncovered lines AND diff coverage is less than total coverage.
-        # To debug, try a step to log outputs like: `echo ${{ toJSON(steps.coverage_comment.outputs) }}`
-        # Once we track base branch, reference_percent_covered will be better to use than new_percent_covered.
-        if: ${{ github.event_name == 'pull_request' && fromJSON(steps.coverage_comment.outputs.diff_total_num_violations) > 0 && steps.coverage_comment.outputs.diff_total_percent_covered < steps.coverage_comment.outputs.new_percent_covered }}
-        run: |
-          echo "Coverage decreased, which is not allowed."
-          echo "Please add some unit tests for the modified code."
-          echo
-          echo "  diff_total_num_violations: ${{ steps.coverage_comment.outputs.diff_total_num_violations }}"
-          echo "  diff_total_percent_covered: ${{ steps.coverage_comment.outputs.diff_total_percent_covered}}"
-          echo "  new_percent_covered: ${{ steps.coverage_comment.outputs.new_percent_covered}}"
-          exit 1

CNAME

@@ -0,0 +1 @@
+docs.all-hands.dev

Development.md

@@ -159,7 +159,7 @@ poetry run pytest ./tests/unit/test_*.py
 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/openhands/runtime:0.59-nikolaik`
+Example: `export SANDBOX_RUNTIME_CONTAINER_IMAGE=ghcr.io/openhands/runtime:0.61-nikolaik`
 
 ## Develop inside Docker container
 

README.md

@@ -51,7 +51,7 @@ Learn more at [docs.all-hands.dev](https://docs.all-hands.dev), or [sign up for
 
 ## ☁️ OpenHands Cloud
 The easiest way to get started with OpenHands is on [OpenHands Cloud](https://app.all-hands.dev),
-which comes with $20 in free credits for new users.
+which comes with $10 in free credits for new users.
 
 ## 💻 Running OpenHands Locally
 
@@ -66,10 +66,10 @@ See the [uv installation guide](https://docs.astral.sh/uv/getting-started/instal
 **Launch OpenHands**:
 ```bash
 # Launch the GUI server
-uvx --python 3.12 --from openhands-ai openhands serve
+uvx --python 3.12 openhands serve
 
 # Or launch the CLI
-uvx --python 3.12 --from openhands-ai openhands
+uvx --python 3.12 openhands
 ```
 
 You'll find OpenHands running at [http://localhost:3000](http://localhost:3000) (for GUI mode)!
@@ -82,17 +82,17 @@ You'll find OpenHands running at [http://localhost:3000](http://localhost:3000)
 You can also run OpenHands directly with Docker:
 
 ```bash
-docker pull docker.all-hands.dev/all-hands-ai/runtime:0.59-nikolaik
+docker pull docker.openhands.dev/openhands/runtime:0.61-nikolaik
 
 docker run -it --rm --pull=always \
-    -e SANDBOX_RUNTIME_CONTAINER_IMAGE=docker.all-hands.dev/all-hands-ai/runtime:0.59-nikolaik \
+    -e SANDBOX_RUNTIME_CONTAINER_IMAGE=docker.openhands.dev/openhands/runtime:0.61-nikolaik \
     -e LOG_ALL_EVENTS=true \
     -v /var/run/docker.sock:/var/run/docker.sock \
     -v ~/.openhands:/.openhands \
     -p 3000:3000 \
     --add-host host.docker.internal:host-gateway \
     --name openhands-app \
-    docker.all-hands.dev/all-hands-ai/openhands:0.59
+    docker.openhands.dev/openhands/openhands:0.61
 ```
 
 </details>

containers/dev/compose.yml

@@ -12,7 +12,7 @@ services:
       - SANDBOX_API_HOSTNAME=host.docker.internal
       - DOCKER_HOST_ADDR=host.docker.internal
       #
-      - SANDBOX_RUNTIME_CONTAINER_IMAGE=${SANDBOX_RUNTIME_CONTAINER_IMAGE:-ghcr.io/openhands/runtime:0.59-nikolaik}
+      - SANDBOX_RUNTIME_CONTAINER_IMAGE=${SANDBOX_RUNTIME_CONTAINER_IMAGE:-ghcr.io/openhands/runtime:0.61-nikolaik}
       - SANDBOX_USER_ID=${SANDBOX_USER_ID:-1234}
       - WORKSPACE_MOUNT_PATH=${WORKSPACE_BASE:-$PWD/workspace}
     ports:

docker-compose.yml

@@ -7,7 +7,7 @@ services:
     image: openhands:latest
     container_name: openhands-app-${DATE:-}
     environment:
-      - SANDBOX_RUNTIME_CONTAINER_IMAGE=${SANDBOX_RUNTIME_CONTAINER_IMAGE:-docker.all-hands.dev/openhands/runtime:0.59-nikolaik}
+      - SANDBOX_RUNTIME_CONTAINER_IMAGE=${SANDBOX_RUNTIME_CONTAINER_IMAGE:-docker.openhands.dev/openhands/runtime:0.61-nikolaik}
       #- SANDBOX_USER_ID=${SANDBOX_USER_ID:-1234} # enable this only if you want a specific non-root sandbox user but you will have to manually adjust permissions of ~/.openhands for this user
       - WORKSPACE_MOUNT_PATH=${WORKSPACE_BASE:-$PWD/workspace}
     ports:

enterprise/migrations/versions/080_add_status_and_updated_at_to_callback.py

@@ -0,0 +1,71 @@
+"""add status and updated_at to callback
+
+Revision ID: 080
+Revises: 079
+Create Date: 2025-11-05 00:00:00.000000
+
+"""
+
+from enum import Enum
+from typing import Sequence, Union
+
+import sqlalchemy as sa
+from alembic import op
+
+# revision identifiers, used by Alembic.
+revision: str = '080'
+down_revision: Union[str, None] = '079'
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+class EventCallbackStatus(Enum):
+    ACTIVE = 'ACTIVE'
+    DISABLED = 'DISABLED'
+    COMPLETED = 'COMPLETED'
+    ERROR = 'ERROR'
+
+
+def upgrade() -> None:
+    """Upgrade schema."""
+    status = sa.Enum(EventCallbackStatus, name='eventcallbackstatus')
+    status.create(op.get_bind(), checkfirst=True)
+    op.add_column(
+        'event_callback',
+        sa.Column('status', status, nullable=False, server_default='ACTIVE'),
+    )
+    op.add_column(
+        'event_callback',
+        sa.Column(
+            'updated_at', sa.DateTime, nullable=False, server_default=sa.func.now()
+        ),
+    )
+    op.drop_index('ix_event_callback_result_event_id')
+    op.drop_column('event_callback_result', 'event_id')
+    op.add_column(
+        'event_callback_result', sa.Column('event_id', sa.String, nullable=True)
+    )
+    op.create_index(
+        op.f('ix_event_callback_result_event_id'),
+        'event_callback_result',
+        ['event_id'],
+        unique=False,
+    )
+
+
+def downgrade() -> None:
+    """Downgrade schema."""
+    op.drop_column('event_callback', 'status')
+    op.drop_column('event_callback', 'updated_at')
+    op.drop_index('ix_event_callback_result_event_id')
+    op.drop_column('event_callback_result', 'event_id')
+    op.add_column(
+        'event_callback_result', sa.Column('event_id', sa.UUID, nullable=True)
+    )
+    op.create_index(
+        op.f('ix_event_callback_result_event_id'),
+        'event_callback_result',
+        ['event_id'],
+        unique=False,
+    )
+    op.execute('DROP TYPE eventcallbackstatus')

enterprise/poetry.lock

@@ -201,19 +201,20 @@ files = [
 
 [[package]]
 name = "anthropic"
-version = "0.65.0"
+version = "0.72.0"
 description = "The official Python library for the anthropic API"
 optional = false
 python-versions = ">=3.8"
 groups = ["main"]
 files = [
-    {file = "anthropic-0.65.0-py3-none-any.whl", hash = "sha256:ba9d9f82678046c74ddf5698ca06d9f5b0f599cfac922ab0d5921638eb448d98"},
-    {file = "anthropic-0.65.0.tar.gz", hash = "sha256:6b6b6942574e54342050dfd42b8d856a8366b171daec147df3b80be4722733b9"},
+    {file = "anthropic-0.72.0-py3-none-any.whl", hash = "sha256:0e9f5a7582f038cab8efbb4c959e49ef654a56bfc7ba2da51b5a7b8a84de2e4d"},
+    {file = "anthropic-0.72.0.tar.gz", hash = "sha256:8971fe76dcffc644f74ac3883069beb1527641115ae0d6eb8fa21c1ce4082f7a"},
 ]
 
 [package.dependencies]
 anyio = ">=3.5.0,<5"
 distro = ">=1.7.0,<2"
+docstring-parser = ">=0.15,<1"
 google-auth = {version = ">=2,<3", extras = ["requests"], optional = true, markers = "extra == \"vertex\""}
 httpx = ">=0.25.0,<1"
 jiter = ">=0.4.0,<1"
@@ -222,7 +223,7 @@ sniffio = "*"
 typing-extensions = ">=4.10,<5"
 
 [package.extras]
-aiohttp = ["aiohttp", "httpx-aiohttp (>=0.1.8)"]
+aiohttp = ["aiohttp", "httpx-aiohttp (>=0.1.9)"]
 bedrock = ["boto3 (>=1.28.57)", "botocore (>=1.31.57)"]
 vertex = ["google-auth[requests] (>=2,<3)"]
 
@@ -681,31 +682,34 @@ crt = ["awscrt (==0.27.6)"]
 
 [[package]]
 name = "browser-use"
-version = "0.7.10"
+version = "0.9.5"
 description = "Make websites accessible for AI agents"
 optional = false
 python-versions = "<4.0,>=3.11"
 groups = ["main"]
 files = [
-    {file = "browser_use-0.7.10-py3-none-any.whl", hash = "sha256:669e12571a0c0c4c93e5fd26abf9e2534eb9bacbc510328aedcab795bd8906a9"},
-    {file = "browser_use-0.7.10.tar.gz", hash = "sha256:f93ce59e06906c12d120360dee4aa33d83618ddf7c9a575dd0ac517d2de7ccbc"},
+    {file = "browser_use-0.9.5-py3-none-any.whl", hash = "sha256:4a2e92847204d1ded269026a99cb0cc0e60e38bd2751fa3f58aedd78f00b4e67"},
+    {file = "browser_use-0.9.5.tar.gz", hash = "sha256:f8285fe253b149d01769a7084883b4cf4db351e2f38e26302c157bcbf14a703f"},
 ]
 
 [package.dependencies]
 aiohttp = "3.12.15"
-anthropic = ">=0.58.2,<1.0.0"
+anthropic = ">=0.68.1,<1.0.0"
 anyio = ">=4.9.0"
 authlib = ">=1.6.0"
 bubus = ">=1.5.6"
 cdp-use = ">=1.4.0"
+click = ">=8.1.8"
+cloudpickle = ">=3.1.1"
 google-api-core = ">=2.25.0"
 google-api-python-client = ">=2.174.0"
 google-auth = ">=2.40.3"
 google-auth-oauthlib = ">=1.2.2"
 google-genai = ">=1.29.0,<2.0.0"
 groq = ">=0.30.0"
-html2text = ">=2025.4.15"
 httpx = ">=0.28.1"
+inquirerpy = ">=0.3.4"
+markdownify = ">=1.2.0"
 mcp = ">=1.10.1"
 ollama = ">=0.5.1"
 openai = ">=1.99.2,<2.0.0"
@@ -720,16 +724,20 @@ pypdf = ">=5.7.0"
 python-dotenv = ">=1.0.1"
 reportlab = ">=4.0.0"
 requests = ">=2.32.3"
+rich = ">=14.0.0"
 screeninfo = {version = ">=0.8.1", markers = "platform_system != \"darwin\""}
 typing-extensions = ">=4.12.2"
 uuid7 = ">=0.1.0"
 
 [package.extras]
-all = ["agentmail (>=0.0.53)", "boto3 (>=1.38.45)", "botocore (>=1.37.23)", "click (>=8.1.8)", "imgcat (>=0.6.0)", "langchain-openai (>=0.3.26)", "rich (>=14.0.0)", "textual (>=3.2.0)"]
+all = ["agentmail (==0.0.59)", "boto3 (>=1.38.45)", "botocore (>=1.37.23)", "imgcat (>=0.6.0)", "langchain-openai (>=0.3.26)", "oci (>=2.126.4)", "textual (>=3.2.0)"]
 aws = ["boto3 (>=1.38.45)"]
-cli = ["click (>=8.1.8)", "rich (>=14.0.0)", "textual (>=3.2.0)"]
-eval = ["anyio (>=4.9.0)", "browserbase (==1.4.0)", "datamodel-code-generator (>=0.26.0)", "hyperbrowser (==0.47.0)", "lmnr[all] (==0.7.10)", "psutil (>=7.0.0)"]
-examples = ["agentmail (>=0.0.53)", "botocore (>=1.37.23)", "imgcat (>=0.6.0)", "langchain-openai (>=0.3.26)"]
+cli = ["textual (>=3.2.0)"]
+cli-oci = ["oci (>=2.126.4)", "textual (>=3.2.0)"]
+code = ["matplotlib (>=3.9.0)", "numpy (>=2.3.2)", "pandas (>=2.2.0)", "tabulate (>=0.9.0)"]
+eval = ["anyio (>=4.9.0)", "datamodel-code-generator (>=0.26.0)", "lmnr[all] (==0.7.17)", "psutil (>=7.0.0)"]
+examples = ["agentmail (==0.0.59)", "botocore (>=1.37.23)", "imgcat (>=0.6.0)", "langchain-openai (>=0.3.26)"]
+oci = ["oci (>=2.126.4)"]
 video = ["imageio[ffmpeg] (>=2.37.0)", "numpy (>=2.3.2)"]
 
 [[package]]
@@ -3525,6 +3533,25 @@ files = [
     {file = "iniconfig-2.1.0.tar.gz", hash = "sha256:3abbd2e30b36733fee78f9c7f7308f2d0050e88f0087fd25c2645f63c773e1c7"},
 ]
 
+[[package]]
+name = "inquirerpy"
+version = "0.3.4"
+description = "Python port of Inquirer.js (A collection of common interactive command-line user interfaces)"
+optional = false
+python-versions = ">=3.7,<4.0"
+groups = ["main"]
+files = [
+    {file = "InquirerPy-0.3.4-py3-none-any.whl", hash = "sha256:c65fdfbac1fa00e3ee4fb10679f4d3ed7a012abf4833910e63c295827fe2a7d4"},
+    {file = "InquirerPy-0.3.4.tar.gz", hash = "sha256:89d2ada0111f337483cb41ae31073108b2ec1e618a49d7110b0d7ade89fc197e"},
+]
+
+[package.dependencies]
+pfzy = ">=0.3.1,<0.4.0"
+prompt-toolkit = ">=3.0.1,<4.0.0"
+
+[package.extras]
+docs = ["Sphinx (>=4.1.2,<5.0.0)", "furo (>=2021.8.17-beta.43,<2022.0.0)", "myst-parser (>=0.15.1,<0.16.0)", "sphinx-autobuild (>=2021.3.14,<2022.0.0)", "sphinx-copybutton (>=0.4.0,<0.5.0)"]
+
 [[package]]
 name = "installer"
 version = "0.7.0"
@@ -4580,6 +4607,62 @@ files = [
     {file = "llvmlite-0.44.0.tar.gz", hash = "sha256:07667d66a5d150abed9157ab6c0b9393c9356f229784a4385c02f99e94fc94d4"},
 ]
 
+[[package]]
+name = "lmnr"
+version = "0.7.20"
+description = "Python SDK for Laminar"
+optional = false
+python-versions = "<4,>=3.10"
+groups = ["main"]
+files = [
+    {file = "lmnr-0.7.20-py3-none-any.whl", hash = "sha256:5f9fa7444e6f96c25e097f66484ff29e632bdd1de0e9346948bf5595f4a8af38"},
+    {file = "lmnr-0.7.20.tar.gz", hash = "sha256:1f484cd618db2d71af65f90a0b8b36d20d80dc91a5138b811575c8677bf7c4fd"},
+]
+
+[package.dependencies]
+grpcio = ">=1"
+httpx = ">=0.24.0"
+opentelemetry-api = ">=1.33.0"
+opentelemetry-exporter-otlp-proto-grpc = ">=1.33.0"
+opentelemetry-exporter-otlp-proto-http = ">=1.33.0"
+opentelemetry-instrumentation = ">=0.54b0"
+opentelemetry-instrumentation-threading = ">=0.57b0"
+opentelemetry-sdk = ">=1.33.0"
+opentelemetry-semantic-conventions = ">=0.54b0"
+opentelemetry-semantic-conventions-ai = ">=0.4.13"
+orjson = ">=3.0.0"
+packaging = ">=22.0"
+pydantic = ">=2.0.3,<3.0.0"
+python-dotenv = ">=1.0"
+tenacity = ">=8.0"
+tqdm = ">=4.0"
+
+[package.extras]
+alephalpha = ["opentelemetry-instrumentation-alephalpha (>=0.47.1)"]
+all = ["opentelemetry-instrumentation-alephalpha (>=0.47.1)", "opentelemetry-instrumentation-bedrock (>=0.47.1)", "opentelemetry-instrumentation-chromadb (>=0.47.1)", "opentelemetry-instrumentation-cohere (>=0.47.1)", "opentelemetry-instrumentation-crewai (>=0.47.1)", "opentelemetry-instrumentation-haystack (>=0.47.1)", "opentelemetry-instrumentation-lancedb (>=0.47.1)", "opentelemetry-instrumentation-langchain (>=0.47.1)", "opentelemetry-instrumentation-llamaindex (>=0.47.1)", "opentelemetry-instrumentation-marqo (>=0.47.1)", "opentelemetry-instrumentation-mcp (>=0.47.1)", "opentelemetry-instrumentation-milvus (>=0.47.1)", "opentelemetry-instrumentation-mistralai (>=0.47.1)", "opentelemetry-instrumentation-ollama (>=0.47.1)", "opentelemetry-instrumentation-pinecone (>=0.47.1)", "opentelemetry-instrumentation-qdrant (>=0.47.1)", "opentelemetry-instrumentation-replicate (>=0.47.1)", "opentelemetry-instrumentation-sagemaker (>=0.47.1)", "opentelemetry-instrumentation-together (>=0.47.1)", "opentelemetry-instrumentation-transformers (>=0.47.1)", "opentelemetry-instrumentation-vertexai (>=0.47.1)", "opentelemetry-instrumentation-watsonx (>=0.47.1)", "opentelemetry-instrumentation-weaviate (>=0.47.1)"]
+bedrock = ["opentelemetry-instrumentation-bedrock (>=0.47.1)"]
+chromadb = ["opentelemetry-instrumentation-chromadb (>=0.47.1)"]
+cohere = ["opentelemetry-instrumentation-cohere (>=0.47.1)"]
+crewai = ["opentelemetry-instrumentation-crewai (>=0.47.1)"]
+haystack = ["opentelemetry-instrumentation-haystack (>=0.47.1)"]
+lancedb = ["opentelemetry-instrumentation-lancedb (>=0.47.1)"]
+langchain = ["opentelemetry-instrumentation-langchain (>=0.47.1)"]
+llamaindex = ["opentelemetry-instrumentation-llamaindex (>=0.47.1)"]
+marqo = ["opentelemetry-instrumentation-marqo (>=0.47.1)"]
+mcp = ["opentelemetry-instrumentation-mcp (>=0.47.1)"]
+milvus = ["opentelemetry-instrumentation-milvus (>=0.47.1)"]
+mistralai = ["opentelemetry-instrumentation-mistralai (>=0.47.1)"]
+ollama = ["opentelemetry-instrumentation-ollama (>=0.47.1)"]
+pinecone = ["opentelemetry-instrumentation-pinecone (>=0.47.1)"]
+qdrant = ["opentelemetry-instrumentation-qdrant (>=0.47.1)"]
+replicate = ["opentelemetry-instrumentation-replicate (>=0.47.1)"]
+sagemaker = ["opentelemetry-instrumentation-sagemaker (>=0.47.1)"]
+together = ["opentelemetry-instrumentation-together (>=0.47.1)"]
+transformers = ["opentelemetry-instrumentation-transformers (>=0.47.1)"]
+vertexai = ["opentelemetry-instrumentation-vertexai (>=0.47.1)"]
+watsonx = ["opentelemetry-instrumentation-watsonx (>=0.47.1)"]
+weaviate = ["opentelemetry-instrumentation-weaviate (>=0.47.1)"]
+
 [[package]]
 name = "lxml"
 version = "6.0.1"
@@ -5737,7 +5820,7 @@ llama = ["llama-index (>=0.12.29,<0.13.0)", "llama-index-core (>=0.12.29,<0.13.0
 
 [[package]]
 name = "openhands-agent-server"
-version = "1.0.0a4"
+version = "1.0.0a5"
 description = "OpenHands Agent Server - REST/WebSocket interface for OpenHands AI Agent"
 optional = false
 python-versions = ">=3.12"
@@ -5758,14 +5841,14 @@ wsproto = ">=1.2.0"
 
 [package.source]
 type = "git"
-url = "https://github.com/OpenHands/agent-sdk.git"
-reference = "ce0a71af55dfce101f7419fbdb0116178f01e109"
-resolved_reference = "ce0a71af55dfce101f7419fbdb0116178f01e109"
+url = "https://github.com/OpenHands/software-agent-sdk.git"
+reference = "d5995c31c55e488d4ab0372d292973bc6fad71f1"
+resolved_reference = "d5995c31c55e488d4ab0372d292973bc6fad71f1"
 subdirectory = "openhands-agent-server"
 
 [[package]]
 name = "openhands-ai"
-version = "0.59.0"
+version = "0.0.0-post.5514+7c9e66194"
 description = "OpenHands: Code Less, Make More"
 optional = false
 python-versions = "^3.12,<3.14"
@@ -5801,13 +5884,14 @@ jupyter_kernel_gateway = "*"
 kubernetes = "^33.1.0"
 libtmux = ">=0.46.2"
 litellm = ">=1.74.3, <1.78.0, !=1.64.4, !=1.67.*"
+lmnr = "^0.7.20"
 memory-profiler = "^0.61.0"
 numpy = "*"
 openai = "1.99.9"
 openhands-aci = "0.3.2"
-openhands-agent-server = {git = "https://github.com/OpenHands/agent-sdk.git", rev = "ce0a71af55dfce101f7419fbdb0116178f01e109", subdirectory = "openhands-agent-server"}
-openhands-sdk = {git = "https://github.com/OpenHands/agent-sdk.git", rev = "ce0a71af55dfce101f7419fbdb0116178f01e109", subdirectory = "openhands-sdk"}
-openhands-tools = {git = "https://github.com/OpenHands/agent-sdk.git", rev = "ce0a71af55dfce101f7419fbdb0116178f01e109", subdirectory = "openhands-tools"}
+openhands-agent-server = {git = "https://github.com/OpenHands/software-agent-sdk.git", rev = "d5995c31c55e488d4ab0372d292973bc6fad71f1", subdirectory = "openhands-agent-server"}
+openhands-sdk = {git = "https://github.com/OpenHands/software-agent-sdk.git", rev = "d5995c31c55e488d4ab0372d292973bc6fad71f1", subdirectory = "openhands-sdk"}
+openhands-tools = {git = "https://github.com/OpenHands/software-agent-sdk.git", rev = "d5995c31c55e488d4ab0372d292973bc6fad71f1", subdirectory = "openhands-tools"}
 opentelemetry-api = "^1.33.1"
 opentelemetry-exporter-otlp-proto-grpc = "^1.33.1"
 pathspec = "^0.12.1"
@@ -5863,7 +5947,7 @@ url = ".."
 
 [[package]]
 name = "openhands-sdk"
-version = "1.0.0a4"
+version = "1.0.0a5"
 description = "OpenHands SDK - Core functionality for building AI agents"
 optional = false
 python-versions = ">=3.12"
@@ -5875,6 +5959,7 @@ develop = false
 fastmcp = ">=2.11.3"
 httpx = ">=0.27.0"
 litellm = ">=1.77.7.dev9"
+lmnr = ">=0.7.20"
 pydantic = ">=2.11.7"
 python-frontmatter = ">=1.1.0"
 python-json-logger = ">=3.3.0"
@@ -5886,14 +5971,14 @@ boto3 = ["boto3 (>=1.35.0)"]
 
 [package.source]
 type = "git"
-url = "https://github.com/OpenHands/agent-sdk.git"
-reference = "ce0a71af55dfce101f7419fbdb0116178f01e109"
-resolved_reference = "ce0a71af55dfce101f7419fbdb0116178f01e109"
+url = "https://github.com/OpenHands/software-agent-sdk.git"
+reference = "d5995c31c55e488d4ab0372d292973bc6fad71f1"
+resolved_reference = "d5995c31c55e488d4ab0372d292973bc6fad71f1"
 subdirectory = "openhands-sdk"
 
 [[package]]
 name = "openhands-tools"
-version = "1.0.0a4"
+version = "1.0.0a5"
 description = "OpenHands Tools - Runtime tools for AI agents"
 optional = false
 python-versions = ">=3.12"
@@ -5904,7 +5989,7 @@ develop = false
 [package.dependencies]
 bashlex = ">=0.18"
 binaryornot = ">=0.4.4"
-browser-use = ">=0.7.7"
+browser-use = ">=0.8.0"
 cachetools = "*"
 func-timeout = ">=4.3.5"
 libtmux = ">=0.46.2"
@@ -5913,9 +5998,9 @@ pydantic = ">=2.11.7"
 
 [package.source]
 type = "git"
-url = "https://github.com/OpenHands/agent-sdk.git"
-reference = "ce0a71af55dfce101f7419fbdb0116178f01e109"
-resolved_reference = "ce0a71af55dfce101f7419fbdb0116178f01e109"
+url = "https://github.com/OpenHands/software-agent-sdk.git"
+reference = "d5995c31c55e488d4ab0372d292973bc6fad71f1"
+resolved_reference = "d5995c31c55e488d4ab0372d292973bc6fad71f1"
 subdirectory = "openhands-tools"
 
 [[package]]
@@ -5988,6 +6073,62 @@ opentelemetry-proto = "1.36.0"
 opentelemetry-sdk = ">=1.36.0,<1.37.0"
 typing-extensions = ">=4.6.0"
 
+[[package]]
+name = "opentelemetry-exporter-otlp-proto-http"
+version = "1.36.0"
+description = "OpenTelemetry Collector Protobuf over HTTP Exporter"
+optional = false
+python-versions = ">=3.9"
+groups = ["main"]
+files = [
+    {file = "opentelemetry_exporter_otlp_proto_http-1.36.0-py3-none-any.whl", hash = "sha256:3d769f68e2267e7abe4527f70deb6f598f40be3ea34c6adc35789bea94a32902"},
+    {file = "opentelemetry_exporter_otlp_proto_http-1.36.0.tar.gz", hash = "sha256:dd3637f72f774b9fc9608ab1ac479f8b44d09b6fb5b2f3df68a24ad1da7d356e"},
+]
+
+[package.dependencies]
+googleapis-common-protos = ">=1.52,<2.0"
+opentelemetry-api = ">=1.15,<2.0"
+opentelemetry-exporter-otlp-proto-common = "1.36.0"
+opentelemetry-proto = "1.36.0"
+opentelemetry-sdk = ">=1.36.0,<1.37.0"
+requests = ">=2.7,<3.0"
+typing-extensions = ">=4.5.0"
+
+[[package]]
+name = "opentelemetry-instrumentation"
+version = "0.57b0"
+description = "Instrumentation Tools & Auto Instrumentation for OpenTelemetry Python"
+optional = false
+python-versions = ">=3.9"
+groups = ["main"]
+files = [
+    {file = "opentelemetry_instrumentation-0.57b0-py3-none-any.whl", hash = "sha256:9109280f44882e07cec2850db28210b90600ae9110b42824d196de357cbddf7e"},
+    {file = "opentelemetry_instrumentation-0.57b0.tar.gz", hash = "sha256:f2a30135ba77cdea2b0e1df272f4163c154e978f57214795d72f40befd4fcf05"},
+]
+
+[package.dependencies]
+opentelemetry-api = ">=1.4,<2.0"
+opentelemetry-semantic-conventions = "0.57b0"
+packaging = ">=18.0"
+wrapt = ">=1.0.0,<2.0.0"
+
+[[package]]
+name = "opentelemetry-instrumentation-threading"
+version = "0.57b0"
+description = "Thread context propagation support for OpenTelemetry"
+optional = false
+python-versions = ">=3.9"
+groups = ["main"]
+files = [
+    {file = "opentelemetry_instrumentation_threading-0.57b0-py3-none-any.whl", hash = "sha256:adfd64857c8c78d6111cf80552311e1713bad64272dd81abdd61f07b892a161b"},
+    {file = "opentelemetry_instrumentation_threading-0.57b0.tar.gz", hash = "sha256:06fa4c98d6bfe4670e7532497670ac202db42afa647ff770aedce0e422421c6e"},
+]
+
+[package.dependencies]
+opentelemetry-api = ">=1.12,<2.0"
+opentelemetry-instrumentation = "0.57b0"
+wrapt = ">=1.0.0,<2.0.0"
+
 [[package]]
 name = "opentelemetry-proto"
 version = "1.36.0"
@@ -6036,6 +6177,115 @@ files = [
 opentelemetry-api = "1.36.0"
 typing-extensions = ">=4.5.0"
 
+[[package]]
+name = "opentelemetry-semantic-conventions-ai"
+version = "0.4.13"
+description = "OpenTelemetry Semantic Conventions Extension for Large Language Models"
+optional = false
+python-versions = "<4,>=3.9"
+groups = ["main"]
+files = [
+    {file = "opentelemetry_semantic_conventions_ai-0.4.13-py3-none-any.whl", hash = "sha256:883a30a6bb5deaec0d646912b5f9f6dcbb9f6f72557b73d0f2560bf25d13e2d5"},
+    {file = "opentelemetry_semantic_conventions_ai-0.4.13.tar.gz", hash = "sha256:94efa9fb4ffac18c45f54a3a338ffeb7eedb7e1bb4d147786e77202e159f0036"},
+]
+
+[[package]]
+name = "orjson"
+version = "3.11.4"
+description = "Fast, correct Python JSON library supporting dataclasses, datetimes, and numpy"
+optional = false
+python-versions = ">=3.9"
+groups = ["main"]
+files = [
+    {file = "orjson-3.11.4-cp310-cp310-macosx_10_15_x86_64.macosx_11_0_arm64.macosx_10_15_universal2.whl", hash = "sha256:e3aa2118a3ece0d25489cbe48498de8a5d580e42e8d9979f65bf47900a15aba1"},
+    {file = "orjson-3.11.4-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:a69ab657a4e6733133a3dca82768f2f8b884043714e8d2b9ba9f52b6efef5c44"},
+    {file = "orjson-3.11.4-cp310-cp310-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:3740bffd9816fc0326ddc406098a3a8f387e42223f5f455f2a02a9f834ead80c"},
+    {file = "orjson-3.11.4-cp310-cp310-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:65fd2f5730b1bf7f350c6dc896173d3460d235c4be007af73986d7cd9a2acd23"},
+    {file = "orjson-3.11.4-cp310-cp310-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:9fdc3ae730541086158d549c97852e2eea6820665d4faf0f41bf99df41bc11ea"},
+    {file = "orjson-3.11.4-cp310-cp310-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:e10b4d65901da88845516ce9f7f9736f9638d19a1d483b3883dc0182e6e5edba"},
+    {file = "orjson-3.11.4-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:fb6a03a678085f64b97f9d4a9ae69376ce91a3a9e9b56a82b1580d8e1d501aff"},
+    {file = "orjson-3.11.4-cp310-cp310-musllinux_1_2_aarch64.whl", hash = "sha256:2c82e4f0b1c712477317434761fbc28b044c838b6b1240d895607441412371ac"},
+    {file = "orjson-3.11.4-cp310-cp310-musllinux_1_2_armv7l.whl", hash = "sha256:d58c166a18f44cc9e2bad03a327dc2d1a3d2e85b847133cfbafd6bfc6719bd79"},
+    {file = "orjson-3.11.4-cp310-cp310-musllinux_1_2_i686.whl", hash = "sha256:94f206766bf1ea30e1382e4890f763bd1eefddc580e08fec1ccdc20ddd95c827"},
+    {file = "orjson-3.11.4-cp310-cp310-musllinux_1_2_x86_64.whl", hash = "sha256:41bf25fb39a34cf8edb4398818523277ee7096689db352036a9e8437f2f3ee6b"},
+    {file = "orjson-3.11.4-cp310-cp310-win32.whl", hash = "sha256:fa9627eba4e82f99ca6d29bc967f09aba446ee2b5a1ea728949ede73d313f5d3"},
+    {file = "orjson-3.11.4-cp310-cp310-win_amd64.whl", hash = "sha256:23ef7abc7fca96632d8174ac115e668c1e931b8fe4dde586e92a500bf1914dcc"},
+    {file = "orjson-3.11.4-cp311-cp311-macosx_10_15_x86_64.macosx_11_0_arm64.macosx_10_15_universal2.whl", hash = "sha256:5e59d23cd93ada23ec59a96f215139753fbfe3a4d989549bcb390f8c00370b39"},
+    {file = "orjson-3.11.4-cp311-cp311-macosx_15_0_arm64.whl", hash = "sha256:5c3aedecfc1beb988c27c79d52ebefab93b6c3921dbec361167e6559aba2d36d"},
+    {file = "orjson-3.11.4-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:da9e5301f1c2caa2a9a4a303480d79c9ad73560b2e7761de742ab39fe59d9175"},
+    {file = "orjson-3.11.4-cp311-cp311-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:8873812c164a90a79f65368f8f96817e59e35d0cc02786a5356f0e2abed78040"},
+    {file = "orjson-3.11.4-cp311-cp311-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:5d7feb0741ebb15204e748f26c9638e6665a5fa93c37a2c73d64f1669b0ddc63"},
+    {file = "orjson-3.11.4-cp311-cp311-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:01ee5487fefee21e6910da4c2ee9eef005bee568a0879834df86f888d2ffbdd9"},
+    {file = "orjson-3.11.4-cp311-cp311-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:3d40d46f348c0321df01507f92b95a377240c4ec31985225a6668f10e2676f9a"},
+    {file = "orjson-3.11.4-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:95713e5fc8af84d8edc75b785d2386f653b63d62b16d681687746734b4dfc0be"},
+    {file = "orjson-3.11.4-cp311-cp311-musllinux_1_2_aarch64.whl", hash = "sha256:ad73ede24f9083614d6c4ca9a85fe70e33be7bf047ec586ee2363bc7418fe4d7"},
+    {file = "orjson-3.11.4-cp311-cp311-musllinux_1_2_armv7l.whl", hash = "sha256:842289889de515421f3f224ef9c1f1efb199a32d76d8d2ca2706fa8afe749549"},
+    {file = "orjson-3.11.4-cp311-cp311-musllinux_1_2_i686.whl", hash = "sha256:3b2427ed5791619851c52a1261b45c233930977e7de8cf36de05636c708fa905"},
+    {file = "orjson-3.11.4-cp311-cp311-musllinux_1_2_x86_64.whl", hash = "sha256:3c36e524af1d29982e9b190573677ea02781456b2e537d5840e4538a5ec41907"},
+    {file = "orjson-3.11.4-cp311-cp311-win32.whl", hash = "sha256:87255b88756eab4a68ec61837ca754e5d10fa8bc47dc57f75cedfeaec358d54c"},
+    {file = "orjson-3.11.4-cp311-cp311-win_amd64.whl", hash = "sha256:e2d5d5d798aba9a0e1fede8d853fa899ce2cb930ec0857365f700dffc2c7af6a"},
+    {file = "orjson-3.11.4-cp311-cp311-win_arm64.whl", hash = "sha256:6bb6bb41b14c95d4f2702bce9975fda4516f1db48e500102fc4d8119032ff045"},
+    {file = "orjson-3.11.4-cp312-cp312-macosx_10_15_x86_64.macosx_11_0_arm64.macosx_10_15_universal2.whl", hash = "sha256:d4371de39319d05d3f482f372720b841c841b52f5385bd99c61ed69d55d9ab50"},
+    {file = "orjson-3.11.4-cp312-cp312-macosx_15_0_arm64.whl", hash = "sha256:e41fd3b3cac850eaae78232f37325ed7d7436e11c471246b87b2cd294ec94853"},
+    {file = "orjson-3.11.4-cp312-cp312-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:600e0e9ca042878c7fdf189cf1b028fe2c1418cc9195f6cb9824eb6ed99cb938"},
+    {file = "orjson-3.11.4-cp312-cp312-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:7bbf9b333f1568ef5da42bc96e18bf30fd7f8d54e9ae066d711056add508e415"},
+    {file = "orjson-3.11.4-cp312-cp312-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:4806363144bb6e7297b8e95870e78d30a649fdc4e23fc84daa80c8ebd366ce44"},
+    {file = "orjson-3.11.4-cp312-cp312-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:ad355e8308493f527d41154e9053b86a5be892b3b359a5c6d5d95cda23601cb2"},
+    {file = "orjson-3.11.4-cp312-cp312-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:c8a7517482667fb9f0ff1b2f16fe5829296ed7a655d04d68cd9711a4d8a4e708"},
+    {file = "orjson-3.11.4-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:97eb5942c7395a171cbfecc4ef6701fc3c403e762194683772df4c54cfbb2210"},
+    {file = "orjson-3.11.4-cp312-cp312-musllinux_1_2_aarch64.whl", hash = "sha256:149d95d5e018bdd822e3f38c103b1a7c91f88d38a88aada5c4e9b3a73a244241"},
+    {file = "orjson-3.11.4-cp312-cp312-musllinux_1_2_armv7l.whl", hash = "sha256:624f3951181eb46fc47dea3d221554e98784c823e7069edb5dbd0dc826ac909b"},
+    {file = "orjson-3.11.4-cp312-cp312-musllinux_1_2_i686.whl", hash = "sha256:03bfa548cf35e3f8b3a96c4e8e41f753c686ff3d8e182ce275b1751deddab58c"},
+    {file = "orjson-3.11.4-cp312-cp312-musllinux_1_2_x86_64.whl", hash = "sha256:525021896afef44a68148f6ed8a8bf8375553d6066c7f48537657f64823565b9"},
+    {file = "orjson-3.11.4-cp312-cp312-win32.whl", hash = "sha256:b58430396687ce0f7d9eeb3dd47761ca7d8fda8e9eb92b3077a7a353a75efefa"},
+    {file = "orjson-3.11.4-cp312-cp312-win_amd64.whl", hash = "sha256:c6dbf422894e1e3c80a177133c0dda260f81428f9de16d61041949f6a2e5c140"},
+    {file = "orjson-3.11.4-cp312-cp312-win_arm64.whl", hash = "sha256:d38d2bc06d6415852224fcc9c0bfa834c25431e466dc319f0edd56cca81aa96e"},
+    {file = "orjson-3.11.4-cp313-cp313-macosx_10_15_x86_64.macosx_11_0_arm64.macosx_10_15_universal2.whl", hash = "sha256:2d6737d0e616a6e053c8b4acc9eccea6b6cce078533666f32d140e4f85002534"},
+    {file = "orjson-3.11.4-cp313-cp313-macosx_15_0_arm64.whl", hash = "sha256:afb14052690aa328cc118a8e09f07c651d301a72e44920b887c519b313d892ff"},
+    {file = "orjson-3.11.4-cp313-cp313-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:38aa9e65c591febb1b0aed8da4d469eba239d434c218562df179885c94e1a3ad"},
+    {file = "orjson-3.11.4-cp313-cp313-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:f2cf4dfaf9163b0728d061bebc1e08631875c51cd30bf47cb9e3293bfbd7dcd5"},
+    {file = "orjson-3.11.4-cp313-cp313-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:89216ff3dfdde0e4070932e126320a1752c9d9a758d6a32ec54b3b9334991a6a"},
+    {file = "orjson-3.11.4-cp313-cp313-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:9daa26ca8e97fae0ce8aa5d80606ef8f7914e9b129b6b5df9104266f764ce436"},
+    {file = "orjson-3.11.4-cp313-cp313-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:5c8b2769dc31883c44a9cd126560327767f848eb95f99c36c9932f51090bfce9"},
+    {file = "orjson-3.11.4-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:1469d254b9884f984026bd9b0fa5bbab477a4bfe558bba6848086f6d43eb5e73"},
+    {file = "orjson-3.11.4-cp313-cp313-musllinux_1_2_aarch64.whl", hash = "sha256:68e44722541983614e37117209a194e8c3ad07838ccb3127d96863c95ec7f1e0"},
+    {file = "orjson-3.11.4-cp313-cp313-musllinux_1_2_armv7l.whl", hash = "sha256:8e7805fda9672c12be2f22ae124dcd7b03928d6c197544fe12174b86553f3196"},
+    {file = "orjson-3.11.4-cp313-cp313-musllinux_1_2_i686.whl", hash = "sha256:04b69c14615fb4434ab867bf6f38b2d649f6f300af30a6705397e895f7aec67a"},
+    {file = "orjson-3.11.4-cp313-cp313-musllinux_1_2_x86_64.whl", hash = "sha256:639c3735b8ae7f970066930e58cf0ed39a852d417c24acd4a25fc0b3da3c39a6"},
+    {file = "orjson-3.11.4-cp313-cp313-win32.whl", hash = "sha256:6c13879c0d2964335491463302a6ca5ad98105fc5db3565499dcb80b1b4bd839"},
+    {file = "orjson-3.11.4-cp313-cp313-win_amd64.whl", hash = "sha256:09bf242a4af98732db9f9a1ec57ca2604848e16f132e3f72edfd3c5c96de009a"},
+    {file = "orjson-3.11.4-cp313-cp313-win_arm64.whl", hash = "sha256:a85f0adf63319d6c1ba06fb0dbf997fced64a01179cf17939a6caca662bf92de"},
+    {file = "orjson-3.11.4-cp314-cp314-macosx_10_15_x86_64.macosx_11_0_arm64.macosx_10_15_universal2.whl", hash = "sha256:42d43a1f552be1a112af0b21c10a5f553983c2a0938d2bbb8ecd8bc9fb572803"},
+    {file = "orjson-3.11.4-cp314-cp314-macosx_15_0_arm64.whl", hash = "sha256:26a20f3fbc6c7ff2cb8e89c4c5897762c9d88cf37330c6a117312365d6781d54"},
+    {file = "orjson-3.11.4-cp314-cp314-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:6e3f20be9048941c7ffa8fc523ccbd17f82e24df1549d1d1fe9317712d19938e"},
+    {file = "orjson-3.11.4-cp314-cp314-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:aac364c758dc87a52e68e349924d7e4ded348dedff553889e4d9f22f74785316"},
+    {file = "orjson-3.11.4-cp314-cp314-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:d5c54a6d76e3d741dcc3f2707f8eeb9ba2a791d3adbf18f900219b62942803b1"},
+    {file = "orjson-3.11.4-cp314-cp314-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:f28485bdca8617b79d44627f5fb04336897041dfd9fa66d383a49d09d86798bc"},
+    {file = "orjson-3.11.4-cp314-cp314-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:bfc2a484cad3585e4ba61985a6062a4c2ed5c7925db6d39f1fa267c9d166487f"},
+    {file = "orjson-3.11.4-cp314-cp314-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:e34dbd508cb91c54f9c9788923daca129fe5b55c5b4eebe713bf5ed3791280cf"},
+    {file = "orjson-3.11.4-cp314-cp314-musllinux_1_2_aarch64.whl", hash = "sha256:b13c478fa413d4b4ee606ec8e11c3b2e52683a640b006bb586b3041c2ca5f606"},
+    {file = "orjson-3.11.4-cp314-cp314-musllinux_1_2_armv7l.whl", hash = "sha256:724ca721ecc8a831b319dcd72cfa370cc380db0bf94537f08f7edd0a7d4e1780"},
+    {file = "orjson-3.11.4-cp314-cp314-musllinux_1_2_i686.whl", hash = "sha256:977c393f2e44845ce1b540e19a786e9643221b3323dae190668a98672d43fb23"},
+    {file = "orjson-3.11.4-cp314-cp314-musllinux_1_2_x86_64.whl", hash = "sha256:1e539e382cf46edec157ad66b0b0872a90d829a6b71f17cb633d6c160a223155"},
+    {file = "orjson-3.11.4-cp314-cp314-win32.whl", hash = "sha256:d63076d625babab9db5e7836118bdfa086e60f37d8a174194ae720161eb12394"},
+    {file = "orjson-3.11.4-cp314-cp314-win_amd64.whl", hash = "sha256:0a54d6635fa3aaa438ae32e8570b9f0de36f3f6562c308d2a2a452e8b0592db1"},
+    {file = "orjson-3.11.4-cp314-cp314-win_arm64.whl", hash = "sha256:78b999999039db3cf58f6d230f524f04f75f129ba3d1ca2ed121f8657e575d3d"},
+    {file = "orjson-3.11.4-cp39-cp39-macosx_10_15_x86_64.macosx_11_0_arm64.macosx_10_15_universal2.whl", hash = "sha256:405261b0a8c62bcbd8e2931c26fdc08714faf7025f45531541e2b29e544b545b"},
+    {file = "orjson-3.11.4-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:af02ff34059ee9199a3546f123a6ab4c86caf1708c79042caf0820dc290a6d4f"},
+    {file = "orjson-3.11.4-cp39-cp39-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:0b2eba969ea4203c177c7b38b36c69519e6067ee68c34dc37081fac74c796e10"},
+    {file = "orjson-3.11.4-cp39-cp39-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:0baa0ea43cfa5b008a28d3c07705cf3ada40e5d347f0f44994a64b1b7b4b5350"},
+    {file = "orjson-3.11.4-cp39-cp39-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:80fd082f5dcc0e94657c144f1b2a3a6479c44ad50be216cf0c244e567f5eae19"},
+    {file = "orjson-3.11.4-cp39-cp39-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:1e3704d35e47d5bee811fb1cbd8599f0b4009b14d451c4c57be5a7e25eb89a13"},
+    {file = "orjson-3.11.4-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:caa447f2b5356779d914658519c874cf3b7629e99e63391ed519c28c8aea4919"},
+    {file = "orjson-3.11.4-cp39-cp39-musllinux_1_2_aarch64.whl", hash = "sha256:bba5118143373a86f91dadb8df41d9457498226698ebdf8e11cbb54d5b0e802d"},
+    {file = "orjson-3.11.4-cp39-cp39-musllinux_1_2_armv7l.whl", hash = "sha256:622463ab81d19ef3e06868b576551587de8e4d518892d1afab71e0fbc1f9cffc"},
+    {file = "orjson-3.11.4-cp39-cp39-musllinux_1_2_i686.whl", hash = "sha256:3e0a700c4b82144b72946b6629968df9762552ee1344bfdb767fecdd634fbd5a"},
+    {file = "orjson-3.11.4-cp39-cp39-musllinux_1_2_x86_64.whl", hash = "sha256:6e18a5c15e764e5f3fc569b47872450b4bcea24f2a6354c0a0e95ad21045d5a9"},
+    {file = "orjson-3.11.4-cp39-cp39-win32.whl", hash = "sha256:fb1c37c71cad991ef4d89c7a634b5ffb4447dbd7ae3ae13e8f5ee7f1775e7ab1"},
+    {file = "orjson-3.11.4-cp39-cp39-win_amd64.whl", hash = "sha256:e2985ce8b8c42d00492d0ed79f2bd2b6460d00f2fa671dfde4bf2e02f49bf5c6"},
+    {file = "orjson-3.11.4.tar.gz", hash = "sha256:39485f4ab4c9b30a3943cfe99e1a213c4776fb69e8abd68f66b83d5a0b0fdc6d"},
+]
+
 [[package]]
 name = "packaging"
 version = "25.0"
@@ -6252,6 +6502,21 @@ files = [
 [package.dependencies]
 ptyprocess = ">=0.5"
 
+[[package]]
+name = "pfzy"
+version = "0.3.4"
+description = "Python port of the fzy fuzzy string matching algorithm"
+optional = false
+python-versions = ">=3.7,<4.0"
+groups = ["main"]
+files = [
+    {file = "pfzy-0.3.4-py3-none-any.whl", hash = "sha256:5f50d5b2b3207fa72e7ec0ef08372ef652685470974a107d0d4999fc5a903a96"},
+    {file = "pfzy-0.3.4.tar.gz", hash = "sha256:717ea765dd10b63618e7298b2d98efd819e0b30cd5905c9707223dceeb94b3f1"},
+]
+
+[package.extras]
+docs = ["Sphinx (>=4.1.2,<5.0.0)", "furo (>=2021.8.17-beta.43,<2022.0.0)", "myst-parser (>=0.15.1,<0.16.0)", "sphinx-autobuild (>=2021.3.14,<2022.0.0)", "sphinx-copybutton (>=0.4.0,<0.5.0)"]
+
 [[package]]
 name = "pg8000"
 version = "1.31.5"

enterprise/server/constants.py

@@ -50,7 +50,7 @@ SUBSCRIPTION_PRICE_DATA = {
     },
 }
 
-DEFAULT_INITIAL_BUDGET = float(os.environ.get('DEFAULT_INITIAL_BUDGET', '20'))
+DEFAULT_INITIAL_BUDGET = float(os.environ.get('DEFAULT_INITIAL_BUDGET', '10'))
 STRIPE_API_KEY = os.environ.get('STRIPE_API_KEY', None)
 STRIPE_WEBHOOK_SECRET = os.environ.get('STRIPE_WEBHOOK_SECRET', None)
 REQUIRE_PAYMENT = os.environ.get('REQUIRE_PAYMENT', '0') in ('1', 'true')

enterprise/storage/saas_conversation_store.py

@@ -35,6 +35,7 @@ class SaasConversationStore(ConversationStore):
             session.query(StoredConversationMetadata)
             .filter(StoredConversationMetadata.user_id == self.user_id)
             .filter(StoredConversationMetadata.conversation_id == conversation_id)
+            .filter(StoredConversationMetadata.conversation_version == 'V0')
         )
 
     def _to_external_model(self, conversation_metadata: StoredConversationMetadata):
@@ -123,6 +124,7 @@ class SaasConversationStore(ConversationStore):
                 conversations = (
                     session.query(StoredConversationMetadata)
                     .filter(StoredConversationMetadata.user_id == self.user_id)
+                    .filter(StoredConversationMetadata.conversation_version == 'V0')
                     .order_by(StoredConversationMetadata.created_at.desc())
                     .offset(offset)
                     .limit(limit + 1)

enterprise/tests/unit/test_saas_settings_store.py

@@ -243,7 +243,7 @@ async def test_update_settings_with_litellm_default(
     # Check that the URL and most of the JSON payload match what we expect
     assert call_args['json']['user_email'] == 'testy@tester.com'
     assert call_args['json']['models'] == []
-    assert call_args['json']['max_budget'] == 20.0
+    assert call_args['json']['max_budget'] == 10.0
     assert call_args['json']['user_id'] == 'user-id'
     assert call_args['json']['teams'] == ['test_team']
     assert call_args['json']['auto_create_key'] is True

evaluation/benchmarks/multi_swe_bench/compute_skip_ids.py

@@ -0,0 +1,79 @@
+import argparse
+import fnmatch
+import json
+from collections import Counter
+from pathlib import Path
+
+
+def find_final_reports(base_dir, pattern=None):
+    base_path = Path(base_dir)
+    if not base_path.exists():
+        raise FileNotFoundError(f'Base directory does not exist: {base_dir}')
+
+    # Find all final_report.json files
+    all_reports = list(base_path.rglob('final_report.json'))
+
+    if pattern is None:
+        return all_reports
+
+    # Filter by pattern
+    filtered_reports = []
+    for report in all_reports:
+        # Get relative path from base_dir for matching
+        rel_path = report.relative_to(base_path)
+        if fnmatch.fnmatch(str(rel_path), pattern):
+            filtered_reports.append(report)
+
+    return filtered_reports
+
+
+def collect_resolved_ids(report_files):
+    id_counter = Counter()
+
+    for report_file in report_files:
+        with open(report_file, 'r') as f:
+            data = json.load(f)
+            if 'resolved_ids' not in data:
+                raise KeyError(f"'resolved_ids' key not found in {report_file}")
+            resolved_ids = data['resolved_ids']
+            id_counter.update(resolved_ids)
+
+    return id_counter
+
+
+def get_skip_ids(id_counter, threshold):
+    return [id_str for id_str, count in id_counter.items() if count >= threshold]
+
+
+def main():
+    parser = argparse.ArgumentParser(
+        description='Compute SKIP_IDS from resolved IDs in final_report.json files'
+    )
+    parser.add_argument(
+        'threshold',
+        type=int,
+        help='Minimum number of times an ID must be resolved to be skipped',
+    )
+    parser.add_argument(
+        '--base-dir',
+        default='evaluation/evaluation_outputs/outputs',
+        help='Base directory to search for final_report.json files (default: evaluation/evaluation_outputs/outputs)',
+    )
+    parser.add_argument(
+        '--pattern',
+        default=None,
+        help='Glob pattern to filter paths (e.g., "*Multi-SWE-RL*/**/*gpt*")',
+    )
+
+    args = parser.parse_args()
+    report_files = find_final_reports(args.base_dir, args.pattern)
+    id_counter = collect_resolved_ids(report_files)
+
+    skip_ids = get_skip_ids(id_counter, args.threshold)
+    skip_ids = [s.replace('/', '__').replace(':pr-', '-') for s in skip_ids]
+    skip_ids = ','.join(sorted(skip_ids))
+    print(skip_ids)
+
+
+if __name__ == '__main__':
+    main()

evaluation/benchmarks/multi_swe_bench/run_infer.py

@@ -747,10 +747,14 @@ def filter_dataset(dataset: pd.DataFrame, filter_column: str) -> pd.DataFrame:
                 subset = dataset[dataset[filter_column].isin(selected_ids)]
                 logger.info(f'Retained {subset.shape[0]} tasks after filtering')
                 return subset
-    skip_ids = os.environ.get('SKIP_IDS', '').split(',')
+    skip_ids = [id for id in os.environ.get('SKIP_IDS', '').split(',') if id]
     if len(skip_ids) > 0:
+        logger.info(f'Dataset size before filtering: {dataset.shape[0]} tasks')
         logger.info(f'Filtering {len(skip_ids)} tasks from "SKIP_IDS"...')
-        return dataset[~dataset[filter_column].isin(skip_ids)]
+        logger.info(f'SKIP_IDS:\n{skip_ids}')
+        filtered_dataset = dataset[~dataset[filter_column].isin(skip_ids)]
+        logger.info(f'Dataset size after filtering: {filtered_dataset.shape[0]} tasks')
+        return filtered_dataset
     return dataset
 
 
@@ -768,6 +772,11 @@ if __name__ == '__main__':
         default='test',
         help='split to evaluate on',
     )
+    parser.add_argument(
+        '--filter_dataset_after_sampling',
+        action='store_true',
+        help='if provided, filter dataset after sampling instead of before',
+    )
     args, _ = parser.parse_known_args()
 
     # NOTE: It is preferable to load datasets from huggingface datasets and perform post-processing
@@ -777,10 +786,24 @@ if __name__ == '__main__':
     logger.info(f'Loading dataset {args.dataset} with split {args.split} ')
     dataset = load_dataset('json', data_files=args.dataset)
     dataset = dataset[args.split]
-    swe_bench_tests = filter_dataset(dataset.to_pandas(), 'instance_id')
-    logger.info(
-        f'Loaded dataset {args.dataset} with split {args.split}: {len(swe_bench_tests)} tasks'
-    )
+    swe_bench_tests = dataset.to_pandas()
+
+    # Determine filter strategy based on flag
+    filter_func = None
+    if args.filter_dataset_after_sampling:
+        # Pass filter as callback to apply after sampling
+        def filter_func(df):
+            return filter_dataset(df, 'instance_id')
+
+        logger.info(
+            f'Loaded dataset {args.dataset} with split {args.split}: {len(swe_bench_tests)} tasks (filtering will occur after sampling)'
+        )
+    else:
+        # Apply filter before sampling
+        swe_bench_tests = filter_dataset(swe_bench_tests, 'instance_id')
+        logger.info(
+            f'Loaded dataset {args.dataset} with split {args.split}: {len(swe_bench_tests)} tasks'
+        )
 
     llm_config = None
     if args.llm_config:
@@ -810,7 +833,9 @@ if __name__ == '__main__':
 
     output_file = os.path.join(metadata.eval_output_dir, 'output.jsonl')
     print(f'### OUTPUT FILE: {output_file} ###')
-    instances = prepare_dataset(swe_bench_tests, output_file, args.eval_n_limit)
+    instances = prepare_dataset(
+        swe_bench_tests, output_file, args.eval_n_limit, filter_func=filter_func
+    )
 
     if len(instances) > 0 and not isinstance(
         instances['FAIL_TO_PASS'][instances['FAIL_TO_PASS'].index[0]], str

evaluation/benchmarks/multi_swe_bench/scripts/rollout_multi_swegym.sh

@@ -8,8 +8,14 @@
 MODEL=$1  # eg your llm config name in config.toml (eg: "llm.claude-3-5-sonnet-20241022-t05")
 EXP_NAME=$2 # "train-t05"
 EVAL_DATASET=$3  # path to original dataset (jsonl file)
-N_WORKERS=${4:-64}
-N_RUNS=${5:-1}
+MAX_ITER=$4
+N_WORKERS=${5:-64}
+N_RUNS=${6:-1}
+EVAL_LIMIT=${7:-}
+SKIP_IDS_THRESHOLD=$8
+SKIP_IDS_PATTERN=$9
+INPUT_SKIP_IDS=${10}
+FILTER_DATASET_AFTER_SAMPLING=${11:-}
 
 export EXP_NAME=$EXP_NAME
 # use 2x resources for rollout since some codebases are pretty resource-intensive
@@ -17,6 +23,7 @@ export DEFAULT_RUNTIME_RESOURCE_FACTOR=2
 echo "MODEL: $MODEL"
 echo "EXP_NAME: $EXP_NAME"
 echo "EVAL_DATASET: $EVAL_DATASET"
+echo "INPUT_SKIP_IDS: $INPUT_SKIP_IDS"
 # Generate DATASET path by adding _with_runtime_ before .jsonl extension
 DATASET="${EVAL_DATASET%.jsonl}_with_runtime_.jsonl"  # path to converted dataset
 
@@ -35,9 +42,6 @@ else
     export SANDBOX_REMOTE_RUNTIME_API_URL="https://runtime.eval.all-hands.dev"
 fi
 
-#EVAL_LIMIT=3000
-MAX_ITER=100
-
 
 # ===== Run inference =====
 source "evaluation/utils/version_control.sh"
@@ -69,17 +73,52 @@ function run_eval() {
     --dataset $DATASET \
     --split $SPLIT"
 
+  # Conditionally add filter flag
+  if [ "$FILTER_DATASET_AFTER_SAMPLING" = "true" ]; then
+    COMMAND="$COMMAND --filter_dataset_after_sampling"
+  fi
+
   echo "Running command: $COMMAND"
   if [ -n "$EVAL_LIMIT" ]; then
     echo "EVAL_LIMIT: $EVAL_LIMIT"
     COMMAND="$COMMAND --eval-n-limit $EVAL_LIMIT"
   fi
 
-  # Run the command
   eval $COMMAND
 }
 
 for run_idx in $(seq 1 $N_RUNS); do
+    if [ -n "$SKIP_IDS_THRESHOLD" ]; then
+        echo "Computing SKIP_IDS for run $run_idx..."
+        SKIP_CMD="poetry run python evaluation/benchmarks/multi_swe_bench/compute_skip_ids.py $SKIP_IDS_THRESHOLD"
+        if [ -n "$SKIP_IDS_PATTERN" ]; then
+            SKIP_CMD="$SKIP_CMD --pattern \"$SKIP_IDS_PATTERN\""
+        fi
+        COMPUTED_SKIP_IDS=$(eval $SKIP_CMD)
+        SKIP_STATUS=$?
+        if [ $SKIP_STATUS -ne 0 ]; then
+            echo "ERROR: Skip IDs computation failed with exit code $SKIP_STATUS"
+            exit $SKIP_STATUS
+        fi
+        echo "COMPUTED_SKIP_IDS: $COMPUTED_SKIP_IDS"
+    else
+        echo "SKIP_IDS_THRESHOLD not provided, skipping SKIP_IDS computation"
+        COMPUTED_SKIP_IDS=""
+    fi
+
+    # Concatenate COMPUTED_SKIP_IDS and INPUT_SKIP_IDS
+    if [ -n "$COMPUTED_SKIP_IDS" ] && [ -n "$INPUT_SKIP_IDS" ]; then
+        export SKIP_IDS="${COMPUTED_SKIP_IDS},${INPUT_SKIP_IDS}"
+    elif [ -n "$COMPUTED_SKIP_IDS" ]; then
+        export SKIP_IDS="$COMPUTED_SKIP_IDS"
+    elif [ -n "$INPUT_SKIP_IDS" ]; then
+        export SKIP_IDS="$INPUT_SKIP_IDS"
+    else
+        unset SKIP_IDS
+    fi
+
+    echo "FINAL SKIP_IDS: $SKIP_IDS"
+    echo ""
 
     while true; do
         echo "### Running inference... ###"

evaluation/utils/shared.py

@@ -9,7 +9,7 @@ import time
 import traceback
 from contextlib import contextmanager
 from inspect import signature
-from typing import Any, Awaitable, Callable, TextIO
+from typing import Any, Awaitable, Callable, Optional, TextIO
 
 import pandas as pd
 from pydantic import BaseModel
@@ -222,6 +222,7 @@ def prepare_dataset(
     eval_n_limit: int,
     eval_ids: list[str] | None = None,
     skip_num: int | None = None,
+    filter_func: Optional[Callable[[pd.DataFrame], pd.DataFrame]] = None,
 ):
     assert 'instance_id' in dataset.columns, (
         "Expected 'instance_id' column in the dataset. You should define your own unique identifier for each instance and use it as the 'instance_id' column."
@@ -265,6 +266,12 @@ def prepare_dataset(
             f'Randomly sampling {eval_n_limit} unique instances with random seed 42.'
         )
 
+    if filter_func is not None:
+        dataset = filter_func(dataset)
+        logger.info(
+            f'Applied filter after sampling: {len(dataset)} instances remaining'
+        )
+
     def make_serializable(instance_dict: dict) -> dict:
         import numpy as np
 

frontend/__tests__/components/features/auth-modal.test.tsx

@@ -8,6 +8,13 @@ vi.mock("#/hooks/use-auth-url", () => ({
   useAuthUrl: () => "https://gitlab.com/oauth/authorize",
 }));
 
+// Mock the useTracking hook
+vi.mock("#/hooks/use-tracking", () => ({
+  useTracking: () => ({
+    trackLoginButtonClick: vi.fn(),
+  }),
+}));
+
 describe("AuthModal", () => {
   beforeEach(() => {
     vi.stubGlobal("location", { href: "" });

frontend/__tests__/components/features/conversation/server-status.test.tsx

@@ -13,34 +13,6 @@ vi.mock("#/hooks/use-agent-state", () => ({
   useAgentState: vi.fn(),
 }));
 
-// Mock the custom hooks
-const mockStartConversationMutate = vi.fn();
-const mockStopConversationMutate = vi.fn();
-
-vi.mock("#/hooks/mutation/use-unified-start-conversation", () => ({
-  useUnifiedStartConversation: () => ({
-    mutate: mockStartConversationMutate,
-  }),
-}));
-
-vi.mock("#/hooks/mutation/use-unified-stop-conversation", () => ({
-  useUnifiedStopConversation: () => ({
-    mutate: mockStopConversationMutate,
-  }),
-}));
-
-vi.mock("#/hooks/use-conversation-id", () => ({
-  useConversationId: () => ({
-    conversationId: "test-conversation-id",
-  }),
-}));
-
-vi.mock("#/hooks/use-user-providers", () => ({
-  useUserProviders: () => ({
-    providers: [],
-  }),
-}));
-
 vi.mock("#/hooks/query/use-task-polling", () => ({
   useTaskPolling: () => ({
     isTask: false,
@@ -66,8 +38,12 @@ vi.mock("react-i18next", async () => {
           COMMON$SERVER_STOPPED: "Server Stopped",
           COMMON$ERROR: "Error",
           COMMON$STARTING: "Starting",
+          COMMON$STOPPING: "Stopping...",
           COMMON$STOP_RUNTIME: "Stop Runtime",
           COMMON$START_RUNTIME: "Start Runtime",
+          CONVERSATION$ERROR_STARTING_CONVERSATION:
+            "Error starting conversation",
+          CONVERSATION$READY: "Ready",
         };
         return translations[key] || key;
       },
@@ -79,10 +55,6 @@ vi.mock("react-i18next", async () => {
 });
 
 describe("ServerStatus", () => {
-  // Mock functions for handlers
-  const mockHandleStop = vi.fn();
-  const mockHandleResumeAgent = vi.fn();
-
   // Helper function to mock agent state with specific state
   const mockAgentStore = (agentState: AgentState) => {
     vi.mocked(useAgentState).mockReturnValue({
@@ -94,248 +66,91 @@ describe("ServerStatus", () => {
     vi.clearAllMocks();
   });
 
-  it("should render server status with different conversation statuses", () => {
-    // Mock agent store to return RUNNING state
+  it("should render server status with RUNNING conversation status", () => {
     mockAgentStore(AgentState.RUNNING);
 
-    // Test RUNNING status
-    const { rerender } = renderWithProviders(
-      <ServerStatus
-        conversationStatus="RUNNING"
-        handleStop={mockHandleStop}
-        handleResumeAgent={mockHandleResumeAgent}
-      />,
-    );
-    expect(screen.getByText("Running")).toBeInTheDocument();
+    renderWithProviders(<ServerStatus conversationStatus="RUNNING" />);
 
-    // Test STOPPED status
-    rerender(
-      <ServerStatus
-        conversationStatus="STOPPED"
-        handleStop={mockHandleStop}
-        handleResumeAgent={mockHandleResumeAgent}
-      />,
-    );
+    expect(screen.getByTestId("server-status")).toBeInTheDocument();
+    expect(screen.getByText("Running")).toBeInTheDocument();
+  });
+
+  it("should render server status with STOPPED conversation status", () => {
+    mockAgentStore(AgentState.RUNNING);
+
+    renderWithProviders(<ServerStatus conversationStatus="STOPPED" />);
+
+    expect(screen.getByTestId("server-status")).toBeInTheDocument();
     expect(screen.getByText("Server Stopped")).toBeInTheDocument();
-
-    // Test STARTING status (shows "Running" due to agent state being RUNNING)
-    rerender(
-      <ServerStatus
-        conversationStatus="STARTING"
-        handleStop={mockHandleStop}
-        handleResumeAgent={mockHandleResumeAgent}
-      />,
-    );
-    expect(screen.getByText("Running")).toBeInTheDocument();
-
-    // Test null status (shows "Running" due to agent state being RUNNING)
-    rerender(
-      <ServerStatus
-        conversationStatus={null}
-        handleStop={mockHandleStop}
-        handleResumeAgent={mockHandleResumeAgent}
-      />,
-    );
-    expect(screen.getByText("Running")).toBeInTheDocument();
   });
 
-  it("should show context menu when clicked with RUNNING status", async () => {
-    const user = userEvent.setup();
+  it("should render STARTING status when agent state is LOADING", () => {
+    mockAgentStore(AgentState.LOADING);
 
-    // Mock agent store to return RUNNING state
+    renderWithProviders(<ServerStatus conversationStatus="STARTING" />);
+
+    expect(screen.getByTestId("server-status")).toBeInTheDocument();
+    expect(screen.getByText("Starting")).toBeInTheDocument();
+  });
+
+  it("should render STARTING status when agent state is INIT", () => {
+    mockAgentStore(AgentState.INIT);
+
+    renderWithProviders(<ServerStatus conversationStatus="STARTING" />);
+
+    expect(screen.getByTestId("server-status")).toBeInTheDocument();
+    expect(screen.getByText("Starting")).toBeInTheDocument();
+  });
+
+  it("should render ERROR status when agent state is ERROR", () => {
+    mockAgentStore(AgentState.ERROR);
+
+    renderWithProviders(<ServerStatus conversationStatus="RUNNING" />);
+
+    expect(screen.getByTestId("server-status")).toBeInTheDocument();
+    expect(screen.getByText("Error")).toBeInTheDocument();
+  });
+
+  it("should render STOPPING status when isPausing is true", () => {
     mockAgentStore(AgentState.RUNNING);
 
     renderWithProviders(
-      <ServerStatus
-        conversationStatus="RUNNING"
-        handleStop={mockHandleStop}
-        handleResumeAgent={mockHandleResumeAgent}
-      />,
+      <ServerStatus conversationStatus="RUNNING" isPausing={true} />,
     );
 
-    const statusContainer = screen.getByText("Running").closest("div");
-    expect(statusContainer).toBeInTheDocument();
-
-    await user.click(statusContainer!);
-
-    // Context menu should appear
-    expect(
-      screen.getByTestId("server-status-context-menu"),
-    ).toBeInTheDocument();
-    expect(screen.getByTestId("stop-server-button")).toBeInTheDocument();
-  });
-
-  it("should show context menu when clicked with STOPPED status", async () => {
-    const user = userEvent.setup();
-
-    // Mock agent store to return STOPPED state
-    mockAgentStore(AgentState.STOPPED);
-
-    renderWithProviders(
-      <ServerStatus
-        conversationStatus="STOPPED"
-        handleStop={mockHandleStop}
-        handleResumeAgent={mockHandleResumeAgent}
-      />,
-    );
-
-    const statusContainer = screen.getByText("Server Stopped").closest("div");
-    expect(statusContainer).toBeInTheDocument();
-
-    await user.click(statusContainer!);
-
-    // Context menu should appear
-    expect(
-      screen.getByTestId("server-status-context-menu"),
-    ).toBeInTheDocument();
-    expect(screen.getByTestId("start-server-button")).toBeInTheDocument();
-  });
-
-  it("should not show context menu when clicked with other statuses", async () => {
-    const user = userEvent.setup();
-
-    // Mock agent store to return RUNNING state
-    mockAgentStore(AgentState.RUNNING);
-
-    renderWithProviders(
-      <ServerStatus
-        conversationStatus="STARTING"
-        handleStop={mockHandleStop}
-        handleResumeAgent={mockHandleResumeAgent}
-      />,
-    );
-
-    const statusContainer = screen.getByText("Running").closest("div");
-    expect(statusContainer).toBeInTheDocument();
-
-    await user.click(statusContainer!);
-
-    // Context menu should not appear
-    expect(
-      screen.queryByTestId("server-status-context-menu"),
-    ).not.toBeInTheDocument();
-  });
-
-  it("should call stop conversation mutation when stop server is clicked", async () => {
-    const user = userEvent.setup();
-
-    // Clear previous calls
-    mockHandleStop.mockClear();
-
-    // Mock agent store to return RUNNING state
-    mockAgentStore(AgentState.RUNNING);
-
-    renderWithProviders(
-      <ServerStatus
-        conversationStatus="RUNNING"
-        handleStop={mockHandleStop}
-        handleResumeAgent={mockHandleResumeAgent}
-      />,
-    );
-
-    const statusContainer = screen.getByText("Running").closest("div");
-    await user.click(statusContainer!);
-
-    const stopButton = screen.getByTestId("stop-server-button");
-    await user.click(stopButton);
-
-    expect(mockHandleStop).toHaveBeenCalledTimes(1);
-  });
-
-  it("should call start conversation mutation when start server is clicked", async () => {
-    const user = userEvent.setup();
-
-    // Clear previous calls
-    mockHandleResumeAgent.mockClear();
-
-    // Mock agent store to return STOPPED state
-    mockAgentStore(AgentState.STOPPED);
-
-    renderWithProviders(
-      <ServerStatus
-        conversationStatus="STOPPED"
-        handleStop={mockHandleStop}
-        handleResumeAgent={mockHandleResumeAgent}
-      />,
-    );
-
-    const statusContainer = screen.getByText("Server Stopped").closest("div");
-    await user.click(statusContainer!);
-
-    const startButton = screen.getByTestId("start-server-button");
-    await user.click(startButton);
-
-    expect(mockHandleResumeAgent).toHaveBeenCalledTimes(1);
-  });
-
-  it("should close context menu after stop server action", async () => {
-    const user = userEvent.setup();
-
-    // Mock agent store to return RUNNING state
-    mockAgentStore(AgentState.RUNNING);
-
-    renderWithProviders(
-      <ServerStatus
-        conversationStatus="RUNNING"
-        handleStop={mockHandleStop}
-        handleResumeAgent={mockHandleResumeAgent}
-      />,
-    );
-
-    const statusContainer = screen.getByText("Running").closest("div");
-    await user.click(statusContainer!);
-
-    const stopButton = screen.getByTestId("stop-server-button");
-    await user.click(stopButton);
-
-    // Context menu should be closed (handled by the component)
-    expect(mockHandleStop).toHaveBeenCalledTimes(1);
-  });
-
-  it("should close context menu after start server action", async () => {
-    const user = userEvent.setup();
-
-    // Mock agent store to return STOPPED state
-    mockAgentStore(AgentState.STOPPED);
-
-    renderWithProviders(
-      <ServerStatus
-        conversationStatus="STOPPED"
-        handleStop={mockHandleStop}
-        handleResumeAgent={mockHandleResumeAgent}
-      />,
-    );
-
-    const statusContainer = screen.getByText("Server Stopped").closest("div");
-    await user.click(statusContainer!);
-
-    const startButton = screen.getByTestId("start-server-button");
-    await user.click(startButton);
-
-    // Context menu should be closed
-    expect(
-      screen.queryByTestId("server-status-context-menu"),
-    ).not.toBeInTheDocument();
+    expect(screen.getByTestId("server-status")).toBeInTheDocument();
+    expect(screen.getByText("Stopping...")).toBeInTheDocument();
   });
 
   it("should handle null conversation status", () => {
-    // Mock agent store to return RUNNING state
+    mockAgentStore(AgentState.RUNNING);
+
+    renderWithProviders(<ServerStatus conversationStatus={null} />);
+
+    expect(screen.getByTestId("server-status")).toBeInTheDocument();
+    expect(screen.getByText("Running")).toBeInTheDocument();
+  });
+
+  it("should apply custom className", () => {
     mockAgentStore(AgentState.RUNNING);
 
     renderWithProviders(
-      <ServerStatus
-        conversationStatus={null}
-        handleStop={mockHandleStop}
-        handleResumeAgent={mockHandleResumeAgent}
-      />,
+      <ServerStatus conversationStatus="RUNNING" className="custom-class" />,
     );
 
-    const statusText = screen.getByText("Running");
-    expect(statusText).toBeInTheDocument();
+    const container = screen.getByTestId("server-status");
+    expect(container).toHaveClass("custom-class");
   });
 });
 
 describe("ServerStatusContextMenu", () => {
+  // Helper function to mock agent state with specific state
+  const mockAgentStore = (agentState: AgentState) => {
+    vi.mocked(useAgentState).mockReturnValue({
+      curAgentState: agentState,
+    });
+  };
+
   const defaultProps = {
     onClose: vi.fn(),
     conversationStatus: "RUNNING" as ConversationStatus,
@@ -346,6 +161,8 @@ describe("ServerStatusContextMenu", () => {
   });
 
   it("should render stop server button when status is RUNNING", () => {
+    mockAgentStore(AgentState.RUNNING);
+
     renderWithProviders(
       <ServerStatusContextMenu
         {...defaultProps}
@@ -354,11 +171,14 @@ describe("ServerStatusContextMenu", () => {
       />,
     );
 
+    expect(screen.getByTestId("server-status")).toBeInTheDocument();
     expect(screen.getByTestId("stop-server-button")).toBeInTheDocument();
     expect(screen.getByText("Stop Runtime")).toBeInTheDocument();
   });
 
   it("should render start server button when status is STOPPED", () => {
+    mockAgentStore(AgentState.RUNNING);
+
     renderWithProviders(
       <ServerStatusContextMenu
         {...defaultProps}
@@ -367,11 +187,14 @@ describe("ServerStatusContextMenu", () => {
       />,
     );
 
+    expect(screen.getByTestId("server-status")).toBeInTheDocument();
     expect(screen.getByTestId("start-server-button")).toBeInTheDocument();
     expect(screen.getByText("Start Runtime")).toBeInTheDocument();
   });
 
   it("should not render stop server button when onStopServer is not provided", () => {
+    mockAgentStore(AgentState.RUNNING);
+
     renderWithProviders(
       <ServerStatusContextMenu
         {...defaultProps}
@@ -379,10 +202,13 @@ describe("ServerStatusContextMenu", () => {
       />,
     );
 
+    expect(screen.getByTestId("server-status")).toBeInTheDocument();
     expect(screen.queryByTestId("stop-server-button")).not.toBeInTheDocument();
   });
 
   it("should not render start server button when onStartServer is not provided", () => {
+    mockAgentStore(AgentState.RUNNING);
+
     renderWithProviders(
       <ServerStatusContextMenu
         {...defaultProps}
@@ -390,12 +216,14 @@ describe("ServerStatusContextMenu", () => {
       />,
     );
 
+    expect(screen.getByTestId("server-status")).toBeInTheDocument();
     expect(screen.queryByTestId("start-server-button")).not.toBeInTheDocument();
   });
 
   it("should call onStopServer when stop button is clicked", async () => {
     const user = userEvent.setup();
     const onStopServer = vi.fn();
+    mockAgentStore(AgentState.RUNNING);
 
     renderWithProviders(
       <ServerStatusContextMenu
@@ -414,6 +242,7 @@ describe("ServerStatusContextMenu", () => {
   it("should call onStartServer when start button is clicked", async () => {
     const user = userEvent.setup();
     const onStartServer = vi.fn();
+    mockAgentStore(AgentState.RUNNING);
 
     renderWithProviders(
       <ServerStatusContextMenu
@@ -430,6 +259,8 @@ describe("ServerStatusContextMenu", () => {
   });
 
   it("should render correct text content for stop server button", () => {
+    mockAgentStore(AgentState.RUNNING);
+
     renderWithProviders(
       <ServerStatusContextMenu
         {...defaultProps}
@@ -444,6 +275,8 @@ describe("ServerStatusContextMenu", () => {
   });
 
   it("should render correct text content for start server button", () => {
+    mockAgentStore(AgentState.RUNNING);
+
     renderWithProviders(
       <ServerStatusContextMenu
         {...defaultProps}
@@ -459,6 +292,7 @@ describe("ServerStatusContextMenu", () => {
 
   it("should call onClose when context menu is closed", () => {
     const onClose = vi.fn();
+    mockAgentStore(AgentState.RUNNING);
 
     renderWithProviders(
       <ServerStatusContextMenu
@@ -475,6 +309,8 @@ describe("ServerStatusContextMenu", () => {
   });
 
   it("should not render any buttons for other conversation statuses", () => {
+    mockAgentStore(AgentState.RUNNING);
+
     renderWithProviders(
       <ServerStatusContextMenu
         {...defaultProps}
@@ -482,6 +318,7 @@ describe("ServerStatusContextMenu", () => {
       />,
     );
 
+    expect(screen.getByTestId("server-status")).toBeInTheDocument();
     expect(screen.queryByTestId("stop-server-button")).not.toBeInTheDocument();
     expect(screen.queryByTestId("start-server-button")).not.toBeInTheDocument();
   });

frontend/__tests__/components/features/microagent-management/microagent-management.test.tsx

@@ -21,6 +21,7 @@ const mockUseConfig = vi.fn();
 const mockUseRepositoryMicroagents = vi.fn();
 const mockUseMicroagentManagementConversations = vi.fn();
 const mockUseSearchRepositories = vi.fn();
+const mockUseCreateConversationAndSubscribeMultiple = vi.fn();
 
 vi.mock("#/hooks/use-user-providers", () => ({
   useUserProviders: () => mockUseUserProviders(),
@@ -47,6 +48,17 @@ vi.mock("#/hooks/query/use-search-repositories", () => ({
   useSearchRepositories: () => mockUseSearchRepositories(),
 }));
 
+vi.mock("#/hooks/use-tracking", () => ({
+  useTracking: () => ({
+    trackEvent: vi.fn(),
+  }),
+}));
+
+vi.mock("#/hooks/use-create-conversation-and-subscribe-multiple", () => ({
+  useCreateConversationAndSubscribeMultiple: () =>
+    mockUseCreateConversationAndSubscribeMultiple(),
+}));
+
 describe("MicroagentManagement", () => {
   const RouterStub = createRoutesStub([
     {
@@ -309,6 +321,16 @@ describe("MicroagentManagement", () => {
       isError: false,
     });
 
+    mockUseCreateConversationAndSubscribeMultiple.mockReturnValue({
+      createConversationAndSubscribe: vi.fn(({ onSuccessCallback }) => {
+        // Immediately call the success callback to close the modal
+        if (onSuccessCallback) {
+          onSuccessCallback();
+        }
+      }),
+      isPending: false,
+    });
+
     // Mock the search repositories hook to return repositories with OpenHands suffixes
     const mockSearchResults =
       getRepositoriesWithOpenHandsSuffix(mockRepositories);

frontend/__tests__/components/features/payment/payment-form.test.tsx

@@ -188,172 +188,4 @@ describe("PaymentForm", () => {
       expect(mockMutate).not.toHaveBeenCalled();
     });
   });
-
-  describe("Cancel Subscription", () => {
-    const getSubscriptionAccessSpy = vi.spyOn(
-      BillingService,
-      "getSubscriptionAccess",
-    );
-    const cancelSubscriptionSpy = vi.spyOn(
-      BillingService,
-      "cancelSubscription",
-    );
-
-    beforeEach(() => {
-      // Mock active subscription
-      getSubscriptionAccessSpy.mockResolvedValue({
-        start_at: "2024-01-01T00:00:00Z",
-        end_at: "2024-12-31T23:59:59Z",
-        created_at: "2024-01-01T00:00:00Z",
-      });
-    });
-
-    it("should render cancel subscription button when user has active subscription", async () => {
-      renderPaymentForm();
-
-      await waitFor(() => {
-        const cancelButton = screen.getByTestId("cancel-subscription-button");
-        expect(cancelButton).toBeInTheDocument();
-        expect(cancelButton).toHaveTextContent("PAYMENT$CANCEL_SUBSCRIPTION");
-      });
-    });
-
-    it("should not render cancel subscription button when user has no subscription", async () => {
-      getSubscriptionAccessSpy.mockResolvedValue(null);
-      renderPaymentForm();
-
-      await waitFor(() => {
-        const cancelButton = screen.queryByTestId("cancel-subscription-button");
-        expect(cancelButton).not.toBeInTheDocument();
-      });
-    });
-
-    it("should show confirmation modal when cancel subscription button is clicked", async () => {
-      const user = userEvent.setup();
-      renderPaymentForm();
-
-      const cancelButton = await screen.findByTestId(
-        "cancel-subscription-button",
-      );
-      await user.click(cancelButton);
-
-      // Should show confirmation modal
-      expect(
-        screen.getByTestId("cancel-subscription-modal"),
-      ).toBeInTheDocument();
-      expect(
-        screen.getByText("PAYMENT$CANCEL_SUBSCRIPTION_TITLE"),
-      ).toBeInTheDocument();
-      // The message should be rendered (either with Trans component or regular text)
-      const modalContent = screen.getByTestId("cancel-subscription-modal");
-      expect(modalContent).toBeInTheDocument();
-      expect(screen.getByTestId("confirm-cancel-button")).toBeInTheDocument();
-      expect(screen.getByTestId("modal-cancel-button")).toBeInTheDocument();
-    });
-
-    it("should close modal when cancel button in modal is clicked", async () => {
-      const user = userEvent.setup();
-      renderPaymentForm();
-
-      const cancelButton = await screen.findByTestId(
-        "cancel-subscription-button",
-      );
-      await user.click(cancelButton);
-
-      // Modal should be visible
-      expect(
-        screen.getByTestId("cancel-subscription-modal"),
-      ).toBeInTheDocument();
-
-      // Click cancel in modal
-      const modalCancelButton = screen.getByTestId("modal-cancel-button");
-      await user.click(modalCancelButton);
-
-      // Modal should be closed
-      expect(
-        screen.queryByTestId("cancel-subscription-modal"),
-      ).not.toBeInTheDocument();
-    });
-
-    it("should call cancel subscription API when confirm button is clicked", async () => {
-      const user = userEvent.setup();
-      renderPaymentForm();
-
-      const cancelButton = await screen.findByTestId(
-        "cancel-subscription-button",
-      );
-      await user.click(cancelButton);
-
-      // Click confirm in modal
-      const confirmButton = screen.getByTestId("confirm-cancel-button");
-      await user.click(confirmButton);
-
-      // Should call the cancel subscription API
-      expect(cancelSubscriptionSpy).toHaveBeenCalled();
-    });
-
-    it("should close modal after successful cancellation", async () => {
-      const user = userEvent.setup();
-      cancelSubscriptionSpy.mockResolvedValue({
-        status: "success",
-        message: "Subscription cancelled successfully",
-      });
-      renderPaymentForm();
-
-      const cancelButton = await screen.findByTestId(
-        "cancel-subscription-button",
-      );
-      await user.click(cancelButton);
-
-      const confirmButton = screen.getByTestId("confirm-cancel-button");
-      await user.click(confirmButton);
-
-      // Wait for API call to complete and modal to close
-      await waitFor(() => {
-        expect(
-          screen.queryByTestId("cancel-subscription-modal"),
-        ).not.toBeInTheDocument();
-      });
-    });
-
-    it("should show next billing date for active subscription", async () => {
-      // Mock active subscription with end_at as next billing date
-      getSubscriptionAccessSpy.mockResolvedValue({
-        start_at: "2024-01-01T00:00:00Z",
-        end_at: "2025-01-01T00:00:00Z",
-        created_at: "2024-01-01T00:00:00Z",
-        cancelled_at: null,
-        stripe_subscription_id: "sub_123",
-      });
-
-      renderPaymentForm();
-
-      await waitFor(() => {
-        const nextBillingInfo = screen.getByTestId("next-billing-date");
-        expect(nextBillingInfo).toBeInTheDocument();
-        // Check that it contains some date-related content (translation key or actual date)
-        expect(nextBillingInfo).toHaveTextContent(
-          /2025|PAYMENT.*BILLING.*DATE/,
-        );
-      });
-    });
-
-    it("should not show next billing date when subscription is cancelled", async () => {
-      // Mock cancelled subscription
-      getSubscriptionAccessSpy.mockResolvedValue({
-        start_at: "2024-01-01T00:00:00Z",
-        end_at: "2025-01-01T00:00:00Z",
-        created_at: "2024-01-01T00:00:00Z",
-        cancelled_at: "2024-06-15T10:30:00Z",
-        stripe_subscription_id: "sub_123",
-      });
-
-      renderPaymentForm();
-
-      await waitFor(() => {
-        const nextBillingInfo = screen.queryByTestId("next-billing-date");
-        expect(nextBillingInfo).not.toBeInTheDocument();
-      });
-    });
-  });
 });

frontend/__tests__/components/jupyter/jupyter.test.tsx

@@ -1,47 +0,0 @@
-import { render, screen } from "@testing-library/react";
-import { JupyterEditor } from "#/components/features/jupyter/jupyter";
-import { vi, describe, it, expect, beforeEach } from "vitest";
-import { AgentState } from "#/types/agent-state";
-import { useAgentState } from "#/hooks/use-agent-state";
-import { useJupyterStore } from "#/state/jupyter-store";
-
-// Mock the agent state hook
-vi.mock("#/hooks/use-agent-state", () => ({
-  useAgentState: vi.fn(),
-}));
-
-// Mock react-i18next
-vi.mock("react-i18next", () => ({
-  useTranslation: () => ({
-    t: (key: string) => key,
-  }),
-}));
-
-describe("JupyterEditor", () => {
-  beforeEach(() => {
-    // Reset the Zustand store before each test
-    useJupyterStore.setState({
-      cells: Array(20).fill({
-        content: "Test cell content",
-        type: "input",
-        imageUrls: undefined,
-      }),
-    });
-  });
-
-  it("should have a scrollable container", () => {
-    // Mock agent state to return RUNNING state (not in RUNTIME_INACTIVE_STATES)
-    vi.mocked(useAgentState).mockReturnValue({
-      curAgentState: AgentState.RUNNING,
-    });
-
-    render(
-      <div style={{ height: "100vh" }}>
-        <JupyterEditor maxWidth={800} />
-      </div>,
-    );
-
-    const container = screen.getByTestId("jupyter-container");
-    expect(container).toHaveClass("flex-1 overflow-y-auto");
-  });
-});

frontend/__tests__/components/terminal/terminal.test.tsx

@@ -11,6 +11,7 @@ const renderTerminal = (commands: Command[] = []) => {
 };
 
 describe.skip("Terminal", () => {
+  // Terminal is now read-only - no user input functionality
   global.ResizeObserver = vi.fn().mockImplementation(() => ({
     observe: vi.fn(),
     disconnect: vi.fn(),
@@ -21,8 +22,6 @@ describe.skip("Terminal", () => {
     write: vi.fn(),
     writeln: vi.fn(),
     dispose: vi.fn(),
-    onKey: vi.fn(),
-    attachCustomKeyEventHandler: vi.fn(),
     loadAddon: vi.fn(),
   };
 

frontend/__tests__/conversation-websocket-handler.test.tsx

@@ -1,6 +1,7 @@
 import { describe, it, expect, beforeAll, afterAll, afterEach } from "vitest";
 import { screen, waitFor, render, cleanup } from "@testing-library/react";
 import { QueryClient, QueryClientProvider } from "@tanstack/react-query";
+import { http, HttpResponse } from "msw";
 import { useOptimisticUserMessageStore } from "#/stores/optimistic-user-message-store";
 import {
   createMockMessageEvent,
@@ -13,8 +14,12 @@ import {
   OptimisticUserMessageStoreComponent,
   ErrorMessageStoreComponent,
 } from "./helpers/websocket-test-components";
-import { ConversationWebSocketProvider } from "#/contexts/conversation-websocket-context";
+import {
+  ConversationWebSocketProvider,
+  useConversationWebSocket,
+} from "#/contexts/conversation-websocket-context";
 import { conversationWebSocketTestSetup } from "./helpers/msw-websocket-setup";
+import { useEventStore } from "#/stores/use-event-store";
 
 // MSW WebSocket mock setup
 const { wsLink, server: mswServer } = conversationWebSocketTestSetup();
@@ -417,7 +422,206 @@ describe("Conversation WebSocket Handler", () => {
     it.todo("should handle send attempts when disconnected");
   });
 
-  // 8. Terminal I/O Tests (ExecuteBashAction and ExecuteBashObservation)
+  // 8. History Loading State Tests
+  describe("History Loading State", () => {
+    it("should track history loading state using event count from API", async () => {
+      const conversationId = "test-conversation-with-history";
+
+      // Mock the event count API to return 3 events
+      const expectedEventCount = 3;
+
+      // Create 3 mock events to simulate history
+      const mockHistoryEvents = [
+        createMockUserMessageEvent({ id: "history-event-1" }),
+        createMockMessageEvent({ id: "history-event-2" }),
+        createMockMessageEvent({ id: "history-event-3" }),
+      ];
+
+      // Set up MSW to mock both the HTTP API and WebSocket connection
+      mswServer.use(
+        http.get("/api/v1/events/count", ({ request }) => {
+          const url = new URL(request.url);
+          const conversationIdParam = url.searchParams.get(
+            "conversation_id__eq",
+          );
+
+          if (conversationIdParam === conversationId) {
+            return HttpResponse.json(expectedEventCount);
+          }
+
+          return HttpResponse.json(0);
+        }),
+        wsLink.addEventListener("connection", ({ client, server }) => {
+          server.connect();
+          // Send all history events
+          mockHistoryEvents.forEach((event) => {
+            client.send(JSON.stringify(event));
+          });
+        }),
+      );
+
+      // Create a test component that displays loading state
+      const HistoryLoadingComponent = () => {
+        const context = useConversationWebSocket();
+        const { events } = useEventStore();
+
+        return (
+          <div>
+            <div data-testid="is-loading-history">
+              {context?.isLoadingHistory ? "true" : "false"}
+            </div>
+            <div data-testid="events-received">{events.length}</div>
+            <div data-testid="expected-event-count">{expectedEventCount}</div>
+          </div>
+        );
+      };
+
+      // Render with WebSocket context
+      renderWithWebSocketContext(
+        <HistoryLoadingComponent />,
+        conversationId,
+        `http://localhost:3000/api/conversations/${conversationId}`,
+      );
+
+      // Initially should be loading history
+      expect(screen.getByTestId("is-loading-history")).toHaveTextContent("true");
+
+      // Wait for all events to be received
+      await waitFor(() => {
+        expect(screen.getByTestId("events-received")).toHaveTextContent("3");
+      });
+
+      // Once all events are received, loading should be complete
+      await waitFor(() => {
+        expect(screen.getByTestId("is-loading-history")).toHaveTextContent(
+          "false",
+        );
+      });
+    });
+
+    it("should handle empty conversation history", async () => {
+      const conversationId = "test-conversation-empty";
+
+      // Set up MSW to mock both the HTTP API and WebSocket connection
+      mswServer.use(
+        http.get("/api/v1/events/count", ({ request }) => {
+          const url = new URL(request.url);
+          const conversationIdParam = url.searchParams.get(
+            "conversation_id__eq",
+          );
+
+          if (conversationIdParam === conversationId) {
+            return HttpResponse.json(0);
+          }
+
+          return HttpResponse.json(0);
+        }),
+        wsLink.addEventListener("connection", ({ server }) => {
+          server.connect();
+          // No events sent for empty history
+        }),
+      );
+
+      // Create a test component that displays loading state
+      const HistoryLoadingComponent = () => {
+        const context = useConversationWebSocket();
+
+        return (
+          <div>
+            <div data-testid="is-loading-history">
+              {context?.isLoadingHistory ? "true" : "false"}
+            </div>
+          </div>
+        );
+      };
+
+      // Render with WebSocket context
+      renderWithWebSocketContext(
+        <HistoryLoadingComponent />,
+        conversationId,
+        `http://localhost:3000/api/conversations/${conversationId}`,
+      );
+
+      // Should quickly transition from loading to not loading when count is 0
+      await waitFor(() => {
+        expect(screen.getByTestId("is-loading-history")).toHaveTextContent(
+          "false",
+        );
+      });
+    });
+
+    it("should handle history loading with large event count", async () => {
+      const conversationId = "test-conversation-large-history";
+
+      // Create 50 mock events to simulate large history
+      const expectedEventCount = 50;
+      const mockHistoryEvents = Array.from({ length: 50 }, (_, i) =>
+        createMockMessageEvent({ id: `history-event-${i + 1}` }),
+      );
+
+      // Set up MSW to mock both the HTTP API and WebSocket connection
+      mswServer.use(
+        http.get("/api/v1/events/count", ({ request }) => {
+          const url = new URL(request.url);
+          const conversationIdParam = url.searchParams.get(
+            "conversation_id__eq",
+          );
+
+          if (conversationIdParam === conversationId) {
+            return HttpResponse.json(expectedEventCount);
+          }
+
+          return HttpResponse.json(0);
+        }),
+        wsLink.addEventListener("connection", ({ client, server }) => {
+          server.connect();
+          // Send all history events
+          mockHistoryEvents.forEach((event) => {
+            client.send(JSON.stringify(event));
+          });
+        }),
+      );
+
+      // Create a test component that displays loading state
+      const HistoryLoadingComponent = () => {
+        const context = useConversationWebSocket();
+        const { events } = useEventStore();
+
+        return (
+          <div>
+            <div data-testid="is-loading-history">
+              {context?.isLoadingHistory ? "true" : "false"}
+            </div>
+            <div data-testid="events-received">{events.length}</div>
+          </div>
+        );
+      };
+
+      // Render with WebSocket context
+      renderWithWebSocketContext(
+        <HistoryLoadingComponent />,
+        conversationId,
+        `http://localhost:3000/api/conversations/${conversationId}`,
+      );
+
+      // Initially should be loading history
+      expect(screen.getByTestId("is-loading-history")).toHaveTextContent("true");
+
+      // Wait for all events to be received
+      await waitFor(() => {
+        expect(screen.getByTestId("events-received")).toHaveTextContent("50");
+      });
+
+      // Once all events are received, loading should be complete
+      await waitFor(() => {
+        expect(screen.getByTestId("is-loading-history")).toHaveTextContent(
+          "false",
+        );
+      });
+    });
+  });
+
+  // 9. Terminal I/O Tests (ExecuteBashAction and ExecuteBashObservation)
   describe("Terminal I/O Integration", () => {
     it("should append command to store when ExecuteBashAction event is received", async () => {
       const { createMockExecuteBashActionEvent } = await import(

frontend/__tests__/helpers/msw-websocket-setup.ts

@@ -38,8 +38,7 @@ export const createWebSocketTestSetup = (
 /**
  * Standard WebSocket test setup for conversation WebSocket handler tests
  * Updated to use the V1 WebSocket URL pattern: /sockets/events/{conversationId}
+ * Uses a wildcard pattern to match any conversation ID
  */
 export const conversationWebSocketTestSetup = () =>
-  createWebSocketTestSetup(
-    "ws://localhost:3000/sockets/events/test-conversation-default",
-  );
+  createWebSocketTestSetup("ws://localhost:3000/sockets/events/*");

frontend/__tests__/hooks/use-terminal.test.tsx

@@ -35,13 +35,12 @@ function TestTerminalComponent() {
 }
 
 describe("useTerminal", () => {
+  // Terminal is read-only - no longer tests user input functionality
   const mockTerminal = vi.hoisted(() => ({
     loadAddon: vi.fn(),
     open: vi.fn(),
     write: vi.fn(),
     writeln: vi.fn(),
-    onKey: vi.fn(),
-    attachCustomKeyEventHandler: vi.fn(),
     dispose: vi.fn(),
   }));
 

frontend/__tests__/hooks/use-websocket.test.ts

@@ -268,7 +268,7 @@ describe("useWebSocket", () => {
     });
 
     // onError handler should have been called
-    expect(onErrorSpy).toHaveBeenCalledOnce();
+    expect(onErrorSpy).toHaveBeenCalled();
   });
 
   it("should provide sendMessage function to send messages to WebSocket", async () => {

frontend/__tests__/routes/llm-settings.test.tsx

@@ -4,14 +4,12 @@ import { beforeEach, describe, expect, it, vi } from "vitest";
 import { QueryClientProvider, QueryClient } from "@tanstack/react-query";
 import LlmSettingsScreen from "#/routes/llm-settings";
 import SettingsService from "#/settings-service/settings-service.api";
-import OptionService from "#/api/option-service/option-service.api";
 import {
   MOCK_DEFAULT_USER_SETTINGS,
   resetTestHandlersMockSettings,
 } from "#/mocks/handlers";
 import * as AdvancedSettingsUtlls from "#/utils/has-advanced-settings-set";
 import * as ToastHandlers from "#/utils/custom-toast-handlers";
-import BillingService from "#/api/billing-service/billing-service.api";
 
 // Mock react-router hooks
 const mockUseSearchParams = vi.fn();
@@ -25,12 +23,6 @@ vi.mock("#/hooks/query/use-is-authed", () => ({
   useIsAuthed: () => mockUseIsAuthed(),
 }));
 
-// Mock useIsAllHandsSaaSEnvironment hook
-const mockUseIsAllHandsSaaSEnvironment = vi.fn();
-vi.mock("#/hooks/use-is-all-hands-saas-environment", () => ({
-  useIsAllHandsSaaSEnvironment: () => mockUseIsAllHandsSaaSEnvironment(),
-}));
-
 const renderLlmSettingsScreen = () =>
   render(<LlmSettingsScreen />, {
     wrapper: ({ children }) => (
@@ -54,9 +46,6 @@ beforeEach(() => {
 
   // Default mock for useIsAuthed - returns authenticated by default
   mockUseIsAuthed.mockReturnValue({ data: true, isLoading: false });
-
-  // Default mock for useIsAllHandsSaaSEnvironment - returns true for SaaS environment
-  mockUseIsAllHandsSaaSEnvironment.mockReturnValue(true);
 });
 
 describe("Content", () => {
@@ -605,9 +594,14 @@ describe("Form submission", () => {
     renderLlmSettingsScreen();
 
     await screen.findByTestId("llm-settings-screen");
+    // Component automatically shows advanced view when advanced settings exist
+    // Switch to basic view to test clearing advanced settings
     const advancedSwitch = screen.getByTestId("advanced-settings-switch");
     await userEvent.click(advancedSwitch);
 
+    // Now we should be in basic view
+    await screen.findByTestId("llm-settings-form-basic");
+
     const provider = screen.getByTestId("llm-provider-input");
     const model = screen.getByTestId("llm-model-input");
 
@@ -731,405 +725,3 @@ describe("Status toasts", () => {
     });
   });
 });
-
-describe("SaaS mode", () => {
-  describe("SaaS subscription", () => {
-    // Common mock configurations
-    const MOCK_SAAS_CONFIG = {
-      APP_MODE: "saas" as const,
-      GITHUB_CLIENT_ID: "fake-github-client-id",
-      POSTHOG_CLIENT_KEY: "fake-posthog-client-key",
-      FEATURE_FLAGS: {
-        ENABLE_BILLING: true,
-        HIDE_LLM_SETTINGS: false,
-        ENABLE_JIRA: false,
-        ENABLE_JIRA_DC: false,
-        ENABLE_LINEAR: false,
-      },
-    };
-
-    const MOCK_ACTIVE_SUBSCRIPTION = {
-      start_at: "2024-01-01",
-      end_at: "2024-12-31",
-      created_at: "2024-01-01",
-    };
-
-    it("should show upgrade banner and prevent all interactions for unsubscribed SaaS users", async () => {
-      // Mock SaaS mode without subscription
-      const getConfigSpy = vi.spyOn(OptionService, "getConfig");
-      getConfigSpy.mockResolvedValue(MOCK_SAAS_CONFIG);
-
-      // Mock subscription access to return null (no subscription)
-      const getSubscriptionAccessSpy = vi.spyOn(
-        BillingService,
-        "getSubscriptionAccess",
-      );
-      getSubscriptionAccessSpy.mockResolvedValue(null);
-
-      // Mock saveSettings to ensure it's not called
-      const saveSettingsSpy = vi.spyOn(SettingsService, "saveSettings");
-
-      renderLlmSettingsScreen();
-      await screen.findByTestId("llm-settings-screen");
-
-      // Should show upgrade banner
-      expect(screen.getByTestId("upgrade-banner")).toBeInTheDocument();
-
-      // Should have a clickable upgrade button
-      const upgradeButton = screen.getByRole("button", { name: /upgrade/i });
-      expect(upgradeButton).toBeInTheDocument();
-      expect(upgradeButton).not.toBeDisabled();
-
-      // Form should be disabled
-      const form = screen.getByTestId("llm-settings-form-basic");
-      expect(form).toHaveAttribute("aria-disabled", "true");
-
-      // All form inputs should be disabled or non-interactive
-      const providerInput = screen.getByTestId("llm-provider-input");
-      const modelInput = screen.getByTestId("llm-model-input");
-      const apiKeyInput = screen.getByTestId("llm-api-key-input");
-      const advancedSwitch = screen.getByTestId("advanced-settings-switch");
-      const submitButton = screen.getByTestId("submit-button");
-
-      // Inputs should be disabled
-      expect(providerInput).toBeDisabled();
-      expect(modelInput).toBeDisabled();
-      expect(apiKeyInput).toBeDisabled();
-      expect(advancedSwitch).toBeDisabled();
-      expect(submitButton).toBeDisabled();
-
-      // Confirmation mode switch is in advanced view, so it's not visible in basic view
-      expect(
-        screen.queryByTestId("enable-confirmation-mode-switch"),
-      ).not.toBeInTheDocument();
-
-      // Try to interact with inputs - they should not respond
-      await userEvent.click(providerInput);
-      await userEvent.type(apiKeyInput, "test-key");
-
-      // Values should not change
-      expect(apiKeyInput).toHaveValue("");
-
-      // Try to submit form - should not call API
-      await userEvent.click(submitButton);
-      expect(saveSettingsSpy).not.toHaveBeenCalled();
-    });
-
-    it("should call subscription checkout API when upgrade button is clicked", async () => {
-      // Mock SaaS mode without subscription
-      const getConfigSpy = vi.spyOn(OptionService, "getConfig");
-      getConfigSpy.mockResolvedValue(MOCK_SAAS_CONFIG);
-
-      // Mock subscription access to return null (no subscription)
-      const getSubscriptionAccessSpy = vi.spyOn(
-        BillingService,
-        "getSubscriptionAccess",
-      );
-      getSubscriptionAccessSpy.mockResolvedValue(null);
-
-      // Mock the subscription checkout API call
-      const createSubscriptionCheckoutSessionSpy = vi.spyOn(
-        BillingService,
-        "createSubscriptionCheckoutSession",
-      );
-      createSubscriptionCheckoutSessionSpy.mockResolvedValue({});
-
-      renderLlmSettingsScreen();
-      await screen.findByTestId("llm-settings-screen");
-
-      // Click the upgrade button
-      const upgradeButton = screen.getByRole("button", { name: /upgrade/i });
-      await userEvent.click(upgradeButton);
-
-      // Should call the subscription checkout API
-      expect(createSubscriptionCheckoutSessionSpy).toHaveBeenCalled();
-    });
-
-    it("should disable upgrade button for unauthenticated users in SaaS mode", async () => {
-      // Mock SaaS mode without subscription
-      const getConfigSpy = vi.spyOn(OptionService, "getConfig");
-      getConfigSpy.mockResolvedValue(MOCK_SAAS_CONFIG);
-
-      // Mock subscription access to return null (no subscription)
-      const getSubscriptionAccessSpy = vi.spyOn(
-        BillingService,
-        "getSubscriptionAccess",
-      );
-      getSubscriptionAccessSpy.mockResolvedValue(null);
-
-      // Mock subscription checkout API
-      const createSubscriptionCheckoutSessionSpy = vi.spyOn(
-        BillingService,
-        "createSubscriptionCheckoutSession",
-      );
-
-      // Mock authentication to return false (unauthenticated) from the start
-      mockUseIsAuthed.mockReturnValue({ data: false, isLoading: false });
-
-      // Mock settings to return default settings even when unauthenticated
-      // This is necessary because the useSettings hook is disabled when user is not authenticated
-      const getSettingsSpy = vi.spyOn(SettingsService, "getSettings");
-      getSettingsSpy.mockResolvedValue(MOCK_DEFAULT_USER_SETTINGS);
-
-      renderLlmSettingsScreen();
-
-      // Wait for either the settings screen or skeleton to appear
-      await waitFor(() => {
-        const settingsScreen = screen.queryByTestId("llm-settings-screen");
-        const skeleton = screen.queryByTestId("app-settings-skeleton");
-        expect(settingsScreen || skeleton).toBeInTheDocument();
-      });
-
-      // If we get the skeleton, the test scenario isn't valid - skip the rest
-      if (screen.queryByTestId("app-settings-skeleton")) {
-        // For unauthenticated users, the settings don't load, so no upgrade banner is shown
-        // This is the expected behavior - unauthenticated users see a skeleton loading state
-        expect(screen.queryByTestId("upgrade-banner")).not.toBeInTheDocument();
-        return;
-      }
-
-      await screen.findByTestId("llm-settings-screen");
-
-      // Should show upgrade banner
-      expect(screen.getByTestId("upgrade-banner")).toBeInTheDocument();
-
-      // Upgrade button should be disabled for unauthenticated users
-      const upgradeButton = screen.getByRole("button", { name: /upgrade/i });
-      expect(upgradeButton).toBeInTheDocument();
-      expect(upgradeButton).toBeDisabled();
-
-      // Clicking disabled button should not call the API
-      await userEvent.click(upgradeButton);
-      expect(createSubscriptionCheckoutSessionSpy).not.toHaveBeenCalled();
-    });
-
-    it("should not show upgrade banner and allow form interaction for subscribed SaaS users", async () => {
-      // Mock SaaS mode with subscription
-      const getConfigSpy = vi.spyOn(OptionService, "getConfig");
-      getConfigSpy.mockResolvedValue(MOCK_SAAS_CONFIG);
-
-      // Mock subscription access to return active subscription
-      const getSubscriptionAccessSpy = vi.spyOn(
-        BillingService,
-        "getSubscriptionAccess",
-      );
-      getSubscriptionAccessSpy.mockResolvedValue(MOCK_ACTIVE_SUBSCRIPTION);
-
-      renderLlmSettingsScreen();
-      await screen.findByTestId("llm-settings-screen");
-
-      // Wait for subscription data to load
-      await waitFor(() => {
-        expect(getSubscriptionAccessSpy).toHaveBeenCalled();
-      });
-
-      // Should NOT show upgrade banner
-      expect(screen.queryByTestId("upgrade-banner")).not.toBeInTheDocument();
-
-      // Form should NOT be disabled
-      const form = screen.getByTestId("llm-settings-form-basic");
-      expect(form).not.toHaveAttribute("aria-disabled", "true");
-    });
-
-    it("should not call save settings API when making changes in disabled form for unsubscribed users", async () => {
-      // Mock SaaS mode without subscription
-      const getConfigSpy = vi.spyOn(OptionService, "getConfig");
-      getConfigSpy.mockResolvedValue(MOCK_SAAS_CONFIG);
-
-      // Mock subscription access to return null (no subscription)
-      const getSubscriptionAccessSpy = vi.spyOn(
-        BillingService,
-        "getSubscriptionAccess",
-      );
-      getSubscriptionAccessSpy.mockResolvedValue(null);
-
-      // Mock saveSettings to track calls
-      const saveSettingsSpy = vi.spyOn(SettingsService, "saveSettings");
-
-      renderLlmSettingsScreen();
-      await screen.findByTestId("llm-settings-screen");
-
-      // Verify that basic form elements are disabled for unsubscribed users
-      const advancedSwitch = screen.getByTestId("advanced-settings-switch");
-      const submitButton = screen.getByTestId("submit-button");
-
-      expect(advancedSwitch).toBeDisabled();
-      expect(submitButton).toBeDisabled();
-
-      // Confirmation mode switch is in advanced view, which can't be accessed when form is disabled
-      expect(
-        screen.queryByTestId("enable-confirmation-mode-switch"),
-      ).not.toBeInTheDocument();
-
-      // Try to submit the form - button should remain disabled
-      await userEvent.click(submitButton);
-
-      // Should NOT call save settings API for unsubscribed users
-      expect(saveSettingsSpy).not.toHaveBeenCalled();
-    });
-
-    it("should show backdrop overlay for unsubscribed users", async () => {
-      // Mock SaaS mode without subscription
-      const getConfigSpy = vi.spyOn(OptionService, "getConfig");
-      getConfigSpy.mockResolvedValue(MOCK_SAAS_CONFIG);
-
-      // Mock subscription access to return null (no subscription)
-      const getSubscriptionAccessSpy = vi.spyOn(
-        BillingService,
-        "getSubscriptionAccess",
-      );
-      getSubscriptionAccessSpy.mockResolvedValue(null);
-
-      renderLlmSettingsScreen();
-      await screen.findByTestId("llm-settings-screen");
-
-      // Wait for subscription data to load
-      await waitFor(() => {
-        expect(getSubscriptionAccessSpy).toHaveBeenCalled();
-      });
-
-      // Should show upgrade banner
-      expect(screen.getByTestId("upgrade-banner")).toBeInTheDocument();
-
-      // Should show backdrop overlay
-      const backdrop = screen.getByTestId("settings-backdrop");
-      expect(backdrop).toBeInTheDocument();
-    });
-
-    it("should not show backdrop overlay for subscribed users", async () => {
-      // Mock SaaS mode with subscription
-      const getConfigSpy = vi.spyOn(OptionService, "getConfig");
-      getConfigSpy.mockResolvedValue(MOCK_SAAS_CONFIG);
-
-      // Mock subscription access to return active subscription
-      const getSubscriptionAccessSpy = vi.spyOn(
-        BillingService,
-        "getSubscriptionAccess",
-      );
-      getSubscriptionAccessSpy.mockResolvedValue(MOCK_ACTIVE_SUBSCRIPTION);
-
-      renderLlmSettingsScreen();
-      await screen.findByTestId("llm-settings-screen");
-
-      // Wait for subscription data to load
-      await waitFor(() => {
-        expect(getSubscriptionAccessSpy).toHaveBeenCalled();
-      });
-
-      // Should NOT show backdrop overlay
-      expect(screen.queryByTestId("settings-backdrop")).not.toBeInTheDocument();
-    });
-
-    it("should display success toast when redirected back with ?checkout=success parameter", async () => {
-      // Mock SaaS mode
-      const getConfigSpy = vi.spyOn(OptionService, "getConfig");
-      getConfigSpy.mockResolvedValue(MOCK_SAAS_CONFIG);
-
-      // Mock subscription access
-      const getSubscriptionAccessSpy = vi.spyOn(
-        BillingService,
-        "getSubscriptionAccess",
-      );
-      getSubscriptionAccessSpy.mockResolvedValue(MOCK_ACTIVE_SUBSCRIPTION);
-
-      // Mock toast handler
-      const displaySuccessToastSpy = vi.spyOn(
-        ToastHandlers,
-        "displaySuccessToast",
-      );
-
-      // Mock URL search params with ?checkout=success
-      mockUseSearchParams.mockReturnValue([
-        {
-          get: (param: string) => (param === "checkout" ? "success" : null),
-        },
-        vi.fn(),
-      ]);
-
-      // Render component with checkout=success parameter
-      renderLlmSettingsScreen();
-      await screen.findByTestId("llm-settings-screen");
-
-      // Verify success toast is displayed with correct message
-      expect(displaySuccessToastSpy).toHaveBeenCalledWith(
-        "SUBSCRIPTION$SUCCESS",
-      );
-    });
-
-    it("should display error toast when redirected back with ?checkout=cancel parameter", async () => {
-      // Mock SaaS mode
-      const getConfigSpy = vi.spyOn(OptionService, "getConfig");
-      getConfigSpy.mockResolvedValue(MOCK_SAAS_CONFIG);
-
-      // Mock subscription access
-      const getSubscriptionAccessSpy = vi.spyOn(
-        BillingService,
-        "getSubscriptionAccess",
-      );
-      getSubscriptionAccessSpy.mockResolvedValue(MOCK_ACTIVE_SUBSCRIPTION);
-
-      // Mock toast handler
-      const displayErrorToastSpy = vi.spyOn(ToastHandlers, "displayErrorToast");
-
-      // Mock URL search params with ?checkout=cancel
-      mockUseSearchParams.mockReturnValue([
-        {
-          get: (param: string) => (param === "checkout" ? "cancel" : null),
-        },
-        vi.fn(),
-      ]);
-
-      // Render component with checkout=cancel parameter
-      renderLlmSettingsScreen();
-      await screen.findByTestId("llm-settings-screen");
-
-      // Verify error toast is displayed with correct message
-      expect(displayErrorToastSpy).toHaveBeenCalledWith("SUBSCRIPTION$FAILURE");
-    });
-
-    it("should show upgrade banner when subscription is expired or disabled", async () => {
-      // Mock SaaS mode
-      const getConfigSpy = vi.spyOn(OptionService, "getConfig");
-      getConfigSpy.mockResolvedValue(MOCK_SAAS_CONFIG);
-
-      // Mock subscription access to return null (expired/disabled subscriptions return null from backend)
-      // The backend only returns active subscriptions within their validity period
-      const getSubscriptionAccessSpy = vi.spyOn(
-        BillingService,
-        "getSubscriptionAccess",
-      );
-      getSubscriptionAccessSpy.mockResolvedValue(null);
-
-      renderLlmSettingsScreen();
-      await screen.findByTestId("llm-settings-screen");
-
-      // Wait for subscription data to load
-      await waitFor(() => {
-        expect(getSubscriptionAccessSpy).toHaveBeenCalled();
-      });
-
-      // Should show upgrade banner for expired/disabled subscriptions (when API returns null)
-      expect(screen.getByTestId("upgrade-banner")).toBeInTheDocument();
-
-      // Form should be disabled
-      const form = screen.getByTestId("llm-settings-form-basic");
-      expect(form).toHaveAttribute("aria-disabled", "true");
-
-      // All form inputs should be disabled
-      const providerInput = screen.getByTestId("llm-provider-input");
-      const modelInput = screen.getByTestId("llm-model-input");
-      const apiKeyInput = screen.getByTestId("llm-api-key-input");
-      const advancedSwitch = screen.getByTestId("advanced-settings-switch");
-
-      expect(providerInput).toBeDisabled();
-      expect(modelInput).toBeDisabled();
-      expect(apiKeyInput).toBeDisabled();
-      expect(advancedSwitch).toBeDisabled();
-
-      // Confirmation mode switch is in advanced view, which can't be accessed when form is disabled
-      expect(
-        screen.queryByTestId("enable-confirmation-mode-switch"),
-      ).not.toBeInTheDocument();
-    });
-  });
-});

frontend/__tests__/services/actions.test.tsx

@@ -5,7 +5,6 @@ import { ActionMessage } from "#/types/message";
 // Mock the store and actions
 const mockDispatch = vi.fn();
 const mockAppendInput = vi.fn();
-const mockAppendJupyterInput = vi.fn();
 
 vi.mock("#/store", () => ({
   default: {
@@ -21,14 +20,6 @@ vi.mock("#/state/command-store", () => ({
   },
 }));
 
-vi.mock("#/state/jupyter-store", () => ({
-  useJupyterStore: {
-    getState: () => ({
-      appendJupyterInput: mockAppendJupyterInput,
-    }),
-  },
-}));
-
 vi.mock("#/state/metrics-slice", () => ({
   setMetrics: vi.fn(),
 }));
@@ -63,10 +54,9 @@ describe("handleActionMessage", () => {
     // Check that appendInput was called with the command
     expect(mockAppendInput).toHaveBeenCalledWith("ls -la");
     expect(mockDispatch).not.toHaveBeenCalled();
-    expect(mockAppendJupyterInput).not.toHaveBeenCalled();
   });
 
-  it("should handle RUN_IPYTHON actions by adding input to Jupyter", async () => {
+  it("should handle RUN_IPYTHON actions as no-op (Jupyter removed)", async () => {
     const { handleActionMessage } = await import("#/services/actions");
 
     const ipythonAction: ActionMessage = {
@@ -84,10 +74,7 @@ describe("handleActionMessage", () => {
     // Handle the action
     handleActionMessage(ipythonAction);
 
-    // Check that appendJupyterInput was called with the code
-    expect(mockAppendJupyterInput).toHaveBeenCalledWith(
-      "print('Hello from Jupyter!')",
-    );
+    // Jupyter functionality has been removed, so nothing should be called
     expect(mockAppendInput).not.toHaveBeenCalled();
   });
 
@@ -112,6 +99,5 @@ describe("handleActionMessage", () => {
     // Check that nothing was dispatched or called
     expect(mockDispatch).not.toHaveBeenCalled();
     expect(mockAppendInput).not.toHaveBeenCalled();
-    expect(mockAppendJupyterInput).not.toHaveBeenCalled();
   });
 });

frontend/__tests__/utils/status.test.ts

@@ -1,5 +1,5 @@
 import { describe, it, expect } from "vitest";
-import { getStatusCode, getIndicatorColor, IndicatorColor } from "../status";
+import { getStatusCode, getIndicatorColor, IndicatorColor } from "#/utils/status";
 import { AgentState } from "#/types/agent-state";
 import { I18nKey } from "#/i18n/declaration";
 
@@ -87,6 +87,36 @@ describe("getStatusCode", () => {
     // Should return runtime status since no agent state
     expect(result).toBe("STATUS$STARTING_RUNTIME");
   });
+
+  it("should prioritize task ERROR status over websocket CONNECTING state", () => {
+    // Test case: Task has errored but websocket is still trying to connect
+    const result = getStatusCode(
+      { id: "", message: "", type: "info", status_update: true }, // statusMessage
+      "CONNECTING", // webSocketStatus (stuck connecting)
+      null, // conversationStatus
+      null, // runtimeStatus
+      AgentState.LOADING, // agentState
+      "ERROR", // taskStatus (ERROR)
+    );
+
+    // Should return error message, not "Connecting..."
+    expect(result).toBe(I18nKey.AGENT_STATUS$ERROR_OCCURRED);
+  });
+
+  it("should show Connecting when task is working and websocket is connecting", () => {
+    // Test case: Task is in progress and websocket is connecting normally
+    const result = getStatusCode(
+      { id: "", message: "", type: "info", status_update: true }, // statusMessage
+      "CONNECTING", // webSocketStatus
+      null, // conversationStatus
+      null, // runtimeStatus
+      AgentState.LOADING, // agentState
+      "WORKING", // taskStatus (in progress)
+    );
+
+    // Should show connecting message since task hasn't errored
+    expect(result).toBe(I18nKey.CHAT_INTERFACE$CONNECTING);
+  });
 });
 
 describe("getIndicatorColor", () => {

frontend/package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "openhands-frontend",
-  "version": "0.59.0",
+  "version": "0.61.0",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "openhands-frontend",
-      "version": "0.59.0",
+      "version": "0.61.0",
       "dependencies": {
         "@heroui/react": "^2.8.4",
         "@heroui/use-infinite-scroll": "^2.2.11",

frontend/package.json

@@ -1,6 +1,6 @@
 {
   "name": "openhands-frontend",
-  "version": "0.59.0",
+  "version": "0.61.0",
   "private": true,
   "type": "module",
   "engines": {

frontend/src/api/conversation-service/conversation-service.api.ts

@@ -187,7 +187,7 @@ class ConversationService {
   static async getRuntimeId(
     conversationId: string,
   ): Promise<{ runtime_id: string }> {
-    const url = `/api/conversations/${conversationId}/config`;
+    const url = `${this.getConversationUrl(conversationId)}/config`;
     const { data } = await openHands.get<{ runtime_id: string }>(url, {
       headers: this.getConversationHeaders(),
     });

frontend/src/api/conversation-service/v1-conversation-service.api.ts

@@ -3,6 +3,7 @@ import { openHands } from "../open-hands-axios";
 import { ConversationTrigger, GetVSCodeUrlResponse } from "../open-hands.types";
 import { Provider } from "#/types/settings";
 import { buildHttpBaseUrl } from "#/utils/websocket-url";
+import { buildSessionHeaders } from "#/utils/utils";
 import type {
   V1SendMessageRequest,
   V1SendMessageResponse,
@@ -13,21 +14,6 @@ import type {
 } from "./v1-conversation-service.types";
 
 class V1ConversationService {
-  /**
-   * Build headers for V1 API requests that require session authentication
-   * @param sessionApiKey Session API key for authentication
-   * @returns Headers object with X-Session-API-Key if provided
-   */
-  private static buildSessionHeaders(
-    sessionApiKey?: string | null,
-  ): Record<string, string> {
-    const headers: Record<string, string> = {};
-    if (sessionApiKey) {
-      headers["X-Session-API-Key"] = sessionApiKey;
-    }
-    return headers;
-  }
-
   /**
    * Build the full URL for V1 runtime-specific endpoints
    * @param conversationUrl The conversation URL (e.g., "http://localhost:54928/api/conversations/...")
@@ -160,7 +146,7 @@ class V1ConversationService {
     sessionApiKey?: string | null,
   ): Promise<GetVSCodeUrlResponse> {
     const url = this.buildRuntimeUrl(conversationUrl, "/api/vscode/url");
-    const headers = this.buildSessionHeaders(sessionApiKey);
+    const headers = buildSessionHeaders(sessionApiKey);
 
     // V1 API returns {url: '...'} instead of {vscode_url: '...'}
     // Map it to match the expected interface
@@ -188,7 +174,7 @@ class V1ConversationService {
       conversationUrl,
       `/api/conversations/${conversationId}/pause`,
     );
-    const headers = this.buildSessionHeaders(sessionApiKey);
+    const headers = buildSessionHeaders(sessionApiKey);
 
     const { data } = await axios.post<{ success: boolean }>(
       url,
@@ -216,7 +202,7 @@ class V1ConversationService {
       conversationUrl,
       `/api/conversations/${conversationId}/run`,
     );
-    const headers = this.buildSessionHeaders(sessionApiKey);
+    const headers = buildSessionHeaders(sessionApiKey);
 
     const { data } = await axios.post<{ success: boolean }>(
       url,
@@ -226,36 +212,6 @@ class V1ConversationService {
     return data;
   }
 
-  /**
-   * Pause a V1 sandbox
-   * Calls the /api/v1/sandboxes/{id}/pause endpoint
-   *
-   * @param sandboxId The sandbox ID to pause
-   * @returns Success response
-   */
-  static async pauseSandbox(sandboxId: string): Promise<{ success: boolean }> {
-    const { data } = await openHands.post<{ success: boolean }>(
-      `/api/v1/sandboxes/${sandboxId}/pause`,
-      {},
-    );
-    return data;
-  }
-
-  /**
-   * Resume a V1 sandbox
-   * Calls the /api/v1/sandboxes/{id}/resume endpoint
-   *
-   * @param sandboxId The sandbox ID to resume
-   * @returns Success response
-   */
-  static async resumeSandbox(sandboxId: string): Promise<{ success: boolean }> {
-    const { data } = await openHands.post<{ success: boolean }>(
-      `/api/v1/sandboxes/${sandboxId}/resume`,
-      {},
-    );
-    return data;
-  }
-
   /**
    * Batch get V1 app conversations by their IDs
    * Returns null for any missing conversations
@@ -305,7 +261,7 @@ class V1ConversationService {
       conversationUrl,
       `/api/file/upload/${encodedPath}`,
     );
-    const headers = this.buildSessionHeaders(sessionApiKey);
+    const headers = buildSessionHeaders(sessionApiKey);
 
     // Create FormData with the file
     const formData = new FormData();
@@ -319,6 +275,19 @@ class V1ConversationService {
       },
     });
   }
+
+  /**
+   * Get the conversation config (runtime_id) for a V1 conversation
+   * @param conversationId The conversation ID
+   * @returns Object containing runtime_id
+   */
+  static async getConversationConfig(
+    conversationId: string,
+  ): Promise<{ runtime_id: string }> {
+    const url = `/api/conversations/${conversationId}/config`;
+    const { data } = await openHands.get<{ runtime_id: string }>(url);
+    return data;
+  }
 }
 
 export default V1ConversationService;

frontend/src/api/conversation-service/v1-conversation-service.types.ts

@@ -1,5 +1,6 @@
 import { ConversationTrigger } from "../open-hands.types";
 import { Provider } from "#/types/settings";
+import { V1SandboxStatus } from "../sandbox-service/sandbox-service.types";
 
 // V1 API Types for requests
 // Note: This represents the serialized API format, not the internal TextContent/ImageContent types
@@ -64,14 +65,7 @@ export interface V1AppConversationStartTaskPage {
   next_page_id: string | null;
 }
 
-export type V1SandboxStatus =
-  | "MISSING"
-  | "STARTING"
-  | "RUNNING"
-  | "STOPPED"
-  | "PAUSED";
-
-export type V1AgentExecutionStatus =
+export type V1ConversationExecutionStatus =
   | "RUNNING"
   | "AWAITING_USER_INPUT"
   | "AWAITING_USER_CONFIRMATION"
@@ -94,7 +88,7 @@ export interface V1AppConversation {
   created_at: string;
   updated_at: string;
   sandbox_status: V1SandboxStatus;
-  agent_status: V1AgentExecutionStatus | null;
+  execution_status: V1ConversationExecutionStatus | null;
   conversation_url: string | null;
   session_api_key: string | null;
 }

frontend/src/api/event-service/event-service.api.ts

@@ -0,0 +1,50 @@
+import axios from "axios";
+import { buildHttpBaseUrl } from "#/utils/websocket-url";
+import { buildSessionHeaders } from "#/utils/utils";
+import type {
+  ConfirmationResponseRequest,
+  ConfirmationResponseResponse,
+} from "./event-service.types";
+import { openHands } from "../open-hands-axios";
+
+class EventService {
+  /**
+   * Respond to a confirmation request in a V1 conversation
+   * @param conversationId The conversation ID
+   * @param conversationUrl The conversation URL (e.g., "http://localhost:54928/api/conversations/...")
+   * @param request The confirmation response request
+   * @param sessionApiKey Session API key for authentication (required for V1)
+   * @returns The confirmation response
+   */
+  static async respondToConfirmation(
+    conversationId: string,
+    conversationUrl: string,
+    request: ConfirmationResponseRequest,
+    sessionApiKey?: string | null,
+  ): Promise<ConfirmationResponseResponse> {
+    // Build the runtime URL using the conversation URL
+    const runtimeUrl = buildHttpBaseUrl(conversationUrl);
+
+    // Build session headers for authentication
+    const headers = buildSessionHeaders(sessionApiKey);
+
+    // Make the API call to the runtime endpoint
+    const { data } = await axios.post<ConfirmationResponseResponse>(
+      `${runtimeUrl}/api/conversations/${conversationId}/events/respond_to_confirmation`,
+      request,
+      { headers },
+    );
+
+    return data;
+  }
+
+  static async getEventCount(conversationId: string): Promise<number> {
+    const params = new URLSearchParams();
+    params.append("conversation_id__eq", conversationId);
+    const { data } = await openHands.get<number>(
+      `/api/v1/events/count?${params.toString()}`,
+    );
+    return data;
+  }
+}
+export default EventService;

frontend/src/api/event-service/event-service.types.ts

@@ -0,0 +1,8 @@
+export interface ConfirmationResponseRequest {
+  accept: boolean;
+  reason?: string;
+}
+
+export interface ConfirmationResponseResponse {
+  success: boolean;
+}

frontend/src/api/git-service/v1-git-service.api.ts

@@ -0,0 +1,89 @@
+import axios from "axios";
+import { buildHttpBaseUrl } from "#/utils/websocket-url";
+import { buildSessionHeaders } from "#/utils/utils";
+import { mapV1ToV0Status } from "#/utils/git-status-mapper";
+import type {
+  GitChange,
+  GitChangeDiff,
+  V1GitChangeStatus,
+} from "../open-hands.types";
+
+interface V1GitChange {
+  status: V1GitChangeStatus;
+  path: string;
+}
+
+class V1GitService {
+  /**
+   * Build the full URL for V1 runtime-specific endpoints
+   * @param conversationUrl The conversation URL (e.g., "http://localhost:54928/api/conversations/...")
+   * @param path The API path (e.g., "/api/git/changes")
+   * @returns Full URL to the runtime endpoint
+   */
+  private static buildRuntimeUrl(
+    conversationUrl: string | null | undefined,
+    path: string,
+  ): string {
+    const baseUrl = buildHttpBaseUrl(conversationUrl);
+    return `${baseUrl}${path}`;
+  }
+
+  /**
+   * Get git changes for a V1 conversation
+   * Uses the agent server endpoint: GET /api/git/changes/{path}
+   * Maps V1 status types (ADDED, DELETED, etc.) to V0 format (A, D, etc.)
+   *
+   * @param conversationUrl The conversation URL (e.g., "http://localhost:54928/api/conversations/...")
+   * @param sessionApiKey Session API key for authentication (required for V1)
+   * @param path The git repository path (e.g., /workspace/project or /workspace/project/OpenHands)
+   * @returns List of git changes with V0-compatible status types
+   */
+  static async getGitChanges(
+    conversationUrl: string | null | undefined,
+    sessionApiKey: string | null | undefined,
+    path: string,
+  ): Promise<GitChange[]> {
+    const encodedPath = encodeURIComponent(path);
+    const url = this.buildRuntimeUrl(
+      conversationUrl,
+      `/api/git/changes/${encodedPath}`,
+    );
+    const headers = buildSessionHeaders(sessionApiKey);
+
+    // V1 API returns V1GitChangeStatus types, we need to map them to V0 format
+    const { data } = await axios.get<V1GitChange[]>(url, { headers });
+
+    // Map V1 statuses to V0 format for compatibility
+    return data.map((change) => ({
+      status: mapV1ToV0Status(change.status),
+      path: change.path,
+    }));
+  }
+
+  /**
+   * Get git change diff for a specific file in a V1 conversation
+   * Uses the agent server endpoint: GET /api/git/diff/{path}
+   *
+   * @param conversationUrl The conversation URL (e.g., "http://localhost:54928/api/conversations/...")
+   * @param sessionApiKey Session API key for authentication (required for V1)
+   * @param path The file path to get diff for
+   * @returns Git change diff
+   */
+  static async getGitChangeDiff(
+    conversationUrl: string | null | undefined,
+    sessionApiKey: string | null | undefined,
+    path: string,
+  ): Promise<GitChangeDiff> {
+    const encodedPath = encodeURIComponent(path);
+    const url = this.buildRuntimeUrl(
+      conversationUrl,
+      `/api/git/diff/${encodedPath}`,
+    );
+    const headers = buildSessionHeaders(sessionApiKey);
+
+    const { data } = await axios.get<GitChangeDiff>(url, { headers });
+    return data;
+  }
+}
+
+export default V1GitService;

frontend/src/api/open-hands.types.ts

@@ -84,8 +84,13 @@ export interface ResultSet<T> {
   next_page_id: string | null;
 }
 
+/**
+ * @deprecated Use V1GitChangeStatus for new code. This type is maintained for backward compatibility with V0 API.
+ */
 export type GitChangeStatus = "M" | "A" | "D" | "R" | "U";
 
+export type V1GitChangeStatus = "MOVED" | "ADDED" | "DELETED" | "UPDATED";
+
 export interface GitChange {
   status: GitChangeStatus;
   path: string;

frontend/src/api/sandbox-service/sandbox-service.api.ts

@@ -0,0 +1,52 @@
+// sandbox-service.api.ts
+// This file contains API methods for /api/v1/sandboxes endpoints.
+
+import { openHands } from "../open-hands-axios";
+import type { V1SandboxInfo } from "./sandbox-service.types";
+
+export class SandboxService {
+  /**
+   * Pause a V1 sandbox
+   * Calls the /api/v1/sandboxes/{id}/pause endpoint
+   */
+  static async pauseSandbox(sandboxId: string): Promise<{ success: boolean }> {
+    const { data } = await openHands.post<{ success: boolean }>(
+      `/api/v1/sandboxes/${sandboxId}/pause`,
+      {},
+    );
+    return data;
+  }
+
+  /**
+   * Resume a V1 sandbox
+   * Calls the /api/v1/sandboxes/{id}/resume endpoint
+   */
+  static async resumeSandbox(sandboxId: string): Promise<{ success: boolean }> {
+    const { data } = await openHands.post<{ success: boolean }>(
+      `/api/v1/sandboxes/${sandboxId}/resume`,
+      {},
+    );
+    return data;
+  }
+
+  /**
+   * Batch get V1 sandboxes by their IDs
+   * Returns null for any missing sandboxes
+   */
+  static async batchGetSandboxes(
+    ids: string[],
+  ): Promise<(V1SandboxInfo | null)[]> {
+    if (ids.length === 0) {
+      return [];
+    }
+    if (ids.length > 100) {
+      throw new Error("Cannot request more than 100 sandboxes at once");
+    }
+    const params = new URLSearchParams();
+    ids.forEach((id) => params.append("id", id));
+    const { data } = await openHands.get<(V1SandboxInfo | null)[]>(
+      `/api/v1/sandboxes?${params.toString()}`,
+    );
+    return data;
+  }
+}

frontend/src/api/sandbox-service/sandbox-service.types.ts

@@ -0,0 +1,24 @@
+// sandbox-service.types.ts
+// This file contains types for Sandbox API.
+
+export type V1SandboxStatus =
+  | "MISSING"
+  | "STARTING"
+  | "RUNNING"
+  | "STOPPED"
+  | "PAUSED";
+
+export interface V1ExposedUrl {
+  name: string;
+  url: string;
+}
+
+export interface V1SandboxInfo {
+  id: string;
+  created_by_user_id: string | null;
+  sandbox_spec_id: string;
+  status: V1SandboxStatus;
+  session_api_key: string | null;
+  exposed_urls: V1ExposedUrl[] | null;
+  created_at: string;
+}

frontend/src/components/features/chat/chat-interface.tsx

@@ -48,6 +48,7 @@ import {
 } from "#/types/v1/type-guards";
 import { useActiveConversation } from "#/hooks/query/use-active-conversation";
 import { useTaskPolling } from "#/hooks/query/use-task-polling";
+import { useConversationWebSocket } from "#/contexts/conversation-websocket-context";
 
 function getEntryPoint(
   hasRepository: boolean | null,
@@ -64,8 +65,10 @@ export function ChatInterface() {
   const { errorMessage } = useErrorMessageStore();
   const { isLoadingMessages } = useWsClient();
   const { isTask } = useTaskPolling();
+  const conversationWebSocket = useConversationWebSocket();
   const { send } = useSendMessage();
   const storeEvents = useEventStore((state) => state.events);
+  const uiEvents = useEventStore((state) => state.uiEvents);
   const { setOptimisticUserMessage, getOptimisticUserMessage } =
     useOptimisticUserMessageStore();
   const { t } = useTranslation();
@@ -94,17 +97,43 @@ export function ChatInterface() {
 
   const isV1Conversation = conversation?.conversation_version === "V1";
 
+  // Track when we should show V1 messages (after DOM has rendered)
+  const [showV1Messages, setShowV1Messages] = React.useState(false);
+  const prevV1LoadingRef = React.useRef(
+    conversationWebSocket?.isLoadingHistory,
+  );
+
+  // Wait for DOM to render before showing V1 messages
+  React.useEffect(() => {
+    const wasLoading = prevV1LoadingRef.current;
+    const isLoading = conversationWebSocket?.isLoadingHistory;
+
+    if (wasLoading && !isLoading) {
+      // Loading just finished - wait for next frame to ensure DOM is ready
+      requestAnimationFrame(() => {
+        setShowV1Messages(true);
+      });
+    } else if (isLoading) {
+      // Reset when loading starts
+      setShowV1Messages(false);
+    }
+
+    prevV1LoadingRef.current = isLoading;
+  }, [conversationWebSocket?.isLoadingHistory]);
+
   // Filter V0 events
   const v0Events = storeEvents
     .filter(isV0Event)
     .filter(isActionOrObservation)
     .filter(shouldRenderEvent);
 
-  // Filter V1 events
-  const v1Events = storeEvents.filter(isV1Event).filter(shouldRenderV1Event);
+  // Filter V1 events - use uiEvents for rendering (actions replaced by observations)
+  const v1UiEvents = uiEvents.filter(isV1Event).filter(shouldRenderV1Event);
+  // Keep full v1 events for lookups (includes both actions and observations)
+  const v1FullEvents = storeEvents.filter(isV1Event);
 
   // Combined events count for tracking
-  const totalEvents = v0Events.length || v1Events.length;
+  const totalEvents = v0Events.length || v1UiEvents.length;
 
   // Check if there are any substantive agent actions (not just system messages)
   const hasSubstantiveAgentActions = React.useMemo(
@@ -202,7 +231,7 @@ export function ChatInterface() {
   };
 
   const v0UserEventsExist = hasUserEvent(v0Events);
-  const v1UserEventsExist = hasV1UserEvent(v1Events);
+  const v1UserEventsExist = hasV1UserEvent(v1FullEvents);
   const userEventsExist = v0UserEventsExist || v1UserEventsExist;
 
   return (
@@ -228,6 +257,14 @@ export function ChatInterface() {
             </div>
           )}
 
+          {(conversationWebSocket?.isLoadingHistory || !showV1Messages) &&
+            isV1Conversation &&
+            !isTask && (
+              <div className="flex justify-center">
+                <LoadingSpinner size="small" />
+              </div>
+            )}
+
           {!isLoadingMessages && v0UserEventsExist && (
             <V0Messages
               messages={v0Events}
@@ -237,13 +274,8 @@ export function ChatInterface() {
             />
           )}
 
-          {v1UserEventsExist && (
-            <V1Messages
-              messages={v1Events}
-              isAwaitingUserConfirmation={
-                curAgentState === AgentState.AWAITING_USER_CONFIRMATION
-              }
-            />
+          {showV1Messages && v1UserEventsExist && (
+            <V1Messages messages={v1UiEvents} allEvents={v1FullEvents} />
           )}
         </div>
 

frontend/src/components/features/chat/components/chat-input-actions.tsx

@@ -1,11 +1,7 @@
-import { ConversationStatus } from "#/types/conversation-status";
-import { ServerStatus } from "#/components/features/controls/server-status";
 import { AgentStatus } from "#/components/features/controls/agent-status";
 import { Tools } from "../../controls/tools";
 import { useUnifiedPauseConversationSandbox } from "#/hooks/mutation/use-unified-stop-conversation";
 import { useConversationId } from "#/hooks/use-conversation-id";
-import { useUnifiedResumeConversationSandbox } from "#/hooks/mutation/use-unified-start-conversation";
-import { useUserProviders } from "#/hooks/use-user-providers";
 import { useActiveConversation } from "#/hooks/query/use-active-conversation";
 import { useSendMessage } from "#/hooks/use-send-message";
 import { generateAgentStateChangeEvent } from "#/services/agent-state-service";
@@ -14,32 +10,23 @@ import { useV1PauseConversation } from "#/hooks/mutation/use-v1-pause-conversati
 import { useV1ResumeConversation } from "#/hooks/mutation/use-v1-resume-conversation";
 
 interface ChatInputActionsProps {
-  conversationStatus: ConversationStatus | null;
   disabled: boolean;
   handleResumeAgent: () => void;
 }
 
 export function ChatInputActions({
-  conversationStatus,
   disabled,
   handleResumeAgent,
 }: ChatInputActionsProps) {
   const { data: conversation } = useActiveConversation();
   const pauseConversationSandboxMutation = useUnifiedPauseConversationSandbox();
-  const resumeConversationSandboxMutation =
-    useUnifiedResumeConversationSandbox();
   const v1PauseConversationMutation = useV1PauseConversation();
   const v1ResumeConversationMutation = useV1ResumeConversation();
   const { conversationId } = useConversationId();
-  const { providers } = useUserProviders();
   const { send } = useSendMessage();
 
   const isV1Conversation = conversation?.conversation_version === "V1";
 
-  const handleStopClick = () => {
-    pauseConversationSandboxMutation.mutate({ conversationId });
-  };
-
   const handlePauseAgent = () => {
     if (isV1Conversation) {
       // V1: Pause the conversation (agent execution)
@@ -62,10 +49,6 @@ export function ChatInputActions({
     handleResumeAgent();
   };
 
-  const handleStartClick = () => {
-    resumeConversationSandboxMutation.mutate({ conversationId, providers });
-  };
-
   const isPausing =
     pauseConversationSandboxMutation.isPending ||
     v1PauseConversationMutation.isPending;
@@ -74,12 +57,6 @@ export function ChatInputActions({
     <div className="w-full flex items-center justify-between">
       <div className="flex items-center gap-1">
         <Tools />
-        <ServerStatus
-          conversationStatus={conversationStatus}
-          isPausing={isPausing}
-          handleStop={handleStopClick}
-          handleResumeAgent={handleStartClick}
-        />
       </div>
       <AgentStatus
         className="ml-2 md:ml-3"

frontend/src/components/features/chat/components/chat-input-container.tsx

@@ -1,9 +1,10 @@
 import React from "react";
-import { ConversationStatus } from "#/types/conversation-status";
 import { DragOver } from "../drag-over";
 import { UploadedFiles } from "../uploaded-files";
 import { ChatInputRow } from "./chat-input-row";
 import { ChatInputActions } from "./chat-input-actions";
+import { useConversationStore } from "#/state/conversation-store";
+import { cn } from "#/utils/utils";
 
 interface ChatInputContainerProps {
   chatContainerRef: React.RefObject<HTMLDivElement | null>;
@@ -11,7 +12,6 @@ interface ChatInputContainerProps {
   disabled: boolean;
   showButton: boolean;
   buttonClassName: string;
-  conversationStatus: ConversationStatus | null;
   chatInputRef: React.RefObject<HTMLDivElement | null>;
   handleFileIconClick: (isDisabled: boolean) => void;
   handleSubmit: () => void;
@@ -32,7 +32,6 @@ export function ChatInputContainer({
   disabled,
   showButton,
   buttonClassName,
-  conversationStatus,
   chatInputRef,
   handleFileIconClick,
   handleSubmit,
@@ -46,10 +45,17 @@ export function ChatInputContainer({
   onFocus,
   onBlur,
 }: ChatInputContainerProps) {
+  const conversationMode = useConversationStore(
+    (state) => state.conversationMode,
+  );
+
   return (
     <div
       ref={chatContainerRef}
-      className="bg-[#25272D] box-border content-stretch flex flex-col items-start justify-center p-4 pt-3 relative rounded-[15px] w-full"
+      className={cn(
+        "bg-[#25272D] box-border content-stretch flex flex-col items-start justify-center p-4 pt-3 relative rounded-[15px] w-full",
+        conversationMode === "plan" && "border border-[#597FF4]",
+      )}
       onDragOver={(e) => onDragOver(e, disabled)}
       onDragLeave={(e) => onDragLeave(e, disabled)}
       onDrop={(e) => onDrop(e, disabled)}
@@ -74,7 +80,6 @@ export function ChatInputContainer({
       />
 
       <ChatInputActions
-        conversationStatus={conversationStatus}
         disabled={disabled}
         handleResumeAgent={handleResumeAgent}
       />

frontend/src/components/features/chat/custom-chat-input.tsx

@@ -137,7 +137,6 @@ export function CustomChatInput({
           disabled={isDisabled}
           showButton={showButton}
           buttonClassName={buttonClassName}
-          conversationStatus={conversationStatus}
           chatInputRef={chatInputRef}
           handleFileIconClick={handleFileIconClick}
           handleSubmit={handleSubmit}

frontend/src/components/features/chat/git-control-bar-pr-button.tsx

@@ -1,10 +1,10 @@
 import { useTranslation } from "react-i18next";
-import posthog from "posthog-js";
 import PRIcon from "#/icons/u-pr.svg?react";
 import { cn, getCreatePRPrompt } from "#/utils/utils";
 import { useUserProviders } from "#/hooks/use-user-providers";
 import { I18nKey } from "#/i18n/declaration";
 import { Provider } from "#/types/settings";
+import { useTracking } from "#/hooks/use-tracking";
 
 interface GitControlBarPrButtonProps {
   onSuggestionsClick: (value: string) => void;
@@ -20,6 +20,7 @@ export function GitControlBarPrButton({
   isConversationReady = true,
 }: GitControlBarPrButtonProps) {
   const { t } = useTranslation();
+  const { trackCreatePrButtonClick } = useTracking();
 
   const { providers } = useUserProviders();
 
@@ -28,7 +29,7 @@ export function GitControlBarPrButton({
     providersAreSet && hasRepository && isConversationReady;
 
   const handlePrClick = () => {
-    posthog.capture("create_pr_button_clicked");
+    trackCreatePrButtonClick();
     onSuggestionsClick(getCreatePRPrompt(currentGitProvider));
   };
 

frontend/src/components/features/chat/git-control-bar-pull-button.tsx

@@ -1,10 +1,10 @@
 import { useTranslation } from "react-i18next";
-import posthog from "posthog-js";
 import ArrowDownIcon from "#/icons/u-arrow-down.svg?react";
 import { cn, getGitPullPrompt } from "#/utils/utils";
 import { useActiveConversation } from "#/hooks/query/use-active-conversation";
 import { useUserProviders } from "#/hooks/use-user-providers";
 import { I18nKey } from "#/i18n/declaration";
+import { useTracking } from "#/hooks/use-tracking";
 
 interface GitControlBarPullButtonProps {
   onSuggestionsClick: (value: string) => void;
@@ -16,6 +16,7 @@ export function GitControlBarPullButton({
   isConversationReady = true,
 }: GitControlBarPullButtonProps) {
   const { t } = useTranslation();
+  const { trackPullButtonClick } = useTracking();
 
   const { data: conversation } = useActiveConversation();
   const { providers } = useUserProviders();
@@ -26,7 +27,7 @@ export function GitControlBarPullButton({
     providersAreSet && hasRepository && isConversationReady;
 
   const handlePullClick = () => {
-    posthog.capture("pull_button_clicked");
+    trackPullButtonClick();
     onSuggestionsClick(getGitPullPrompt());
   };
 

frontend/src/components/features/chat/git-control-bar-push-button.tsx

@@ -1,10 +1,10 @@
 import { useTranslation } from "react-i18next";
-import posthog from "posthog-js";
 import ArrowUpIcon from "#/icons/u-arrow-up.svg?react";
 import { cn, getGitPushPrompt } from "#/utils/utils";
 import { useUserProviders } from "#/hooks/use-user-providers";
 import { I18nKey } from "#/i18n/declaration";
 import { Provider } from "#/types/settings";
+import { useTracking } from "#/hooks/use-tracking";
 
 interface GitControlBarPushButtonProps {
   onSuggestionsClick: (value: string) => void;
@@ -20,6 +20,7 @@ export function GitControlBarPushButton({
   isConversationReady = true,
 }: GitControlBarPushButtonProps) {
   const { t } = useTranslation();
+  const { trackPushButtonClick } = useTracking();
 
   const { providers } = useUserProviders();
 
@@ -28,7 +29,7 @@ export function GitControlBarPushButton({
     providersAreSet && hasRepository && isConversationReady;
 
   const handlePushClick = () => {
-    posthog.capture("push_button_clicked");
+    trackPushButtonClick();
     onSuggestionsClick(getGitPushPrompt(currentGitProvider));
   };
 

frontend/src/components/features/controls/agent-status.tsx

@@ -13,6 +13,7 @@ import { useConversationStore } from "#/state/conversation-store";
 import CircleErrorIcon from "#/icons/circle-error.svg?react";
 import { useAgentState } from "#/hooks/use-agent-state";
 import { useUnifiedWebSocketStatus } from "#/hooks/use-unified-websocket-status";
+import { useTaskPolling } from "#/hooks/query/use-task-polling";
 
 export interface AgentStatusProps {
   className?: string;
@@ -35,6 +36,7 @@ export function AgentStatus({
   const { curStatusMessage } = useStatusStore();
   const webSocketStatus = useUnifiedWebSocketStatus();
   const { data: conversation } = useActiveConversation();
+  const { taskStatus } = useTaskPolling();
 
   const statusCode = getStatusCode(
     curStatusMessage,
@@ -42,25 +44,34 @@ export function AgentStatus({
     conversation?.status || null,
     conversation?.runtime_status || null,
     curAgentState,
+    taskStatus,
   );
 
+  const isTaskLoading =
+    taskStatus && taskStatus !== "ERROR" && taskStatus !== "READY";
+
   const shouldShownAgentLoading =
     isPausing ||
     curAgentState === AgentState.INIT ||
     curAgentState === AgentState.LOADING ||
-    webSocketStatus === "CONNECTING";
+    (webSocketStatus === "CONNECTING" && taskStatus !== "ERROR") ||
+    isTaskLoading;
 
   const shouldShownAgentError =
     curAgentState === AgentState.ERROR ||
-    curAgentState === AgentState.RATE_LIMITED;
+    curAgentState === AgentState.RATE_LIMITED ||
+    webSocketStatus === "DISCONNECTED" ||
+    taskStatus === "ERROR";
 
   const shouldShownAgentStop = curAgentState === AgentState.RUNNING;
 
-  const shouldShownAgentResume = curAgentState === AgentState.STOPPED;
+  const shouldShownAgentResume =
+    curAgentState === AgentState.STOPPED || curAgentState === AgentState.PAUSED;
 
   // Update global state when agent loading condition changes
   useEffect(() => {
-    setShouldShownAgentLoading(shouldShownAgentLoading);
+    if (shouldShownAgentLoading)
+      setShouldShownAgentLoading(shouldShownAgentLoading);
   }, [shouldShownAgentLoading, setShouldShownAgentLoading]);
 
   return (

frontend/src/components/features/controls/server-status-context-menu-icon-text.tsx

@@ -13,7 +13,7 @@ export function ServerStatusContextMenuIconText({
 }: ServerStatusContextMenuIconTextProps) {
   return (
     <button
-      className="flex items-center gap-2 p-2 hover:bg-[#5C5D62] rounded text-sm text-white font-normal leading-5 cursor-pointer w-full"
+      className="flex items-center justify-between p-2 hover:bg-[#5C5D62] rounded text-sm text-white font-normal leading-5 cursor-pointer w-full"
       onClick={onClick}
       data-testid={testId}
       type="button"

frontend/src/components/features/controls/server-status-context-menu.tsx

@@ -6,6 +6,9 @@ import { ConversationStatus } from "#/types/conversation-status";
 import StopCircleIcon from "#/icons/stop-circle.svg?react";
 import PlayCircleIcon from "#/icons/play-circle.svg?react";
 import { ServerStatusContextMenuIconText } from "./server-status-context-menu-icon-text";
+import { ServerStatus } from "./server-status";
+import { Divider } from "#/ui/divider";
+import { cn } from "#/utils/utils";
 
 interface ServerStatusContextMenuProps {
   onClose: () => void;
@@ -13,6 +16,8 @@ interface ServerStatusContextMenuProps {
   onStartServer?: (event: React.MouseEvent<HTMLButtonElement>) => void;
   conversationStatus: ConversationStatus | null;
   position?: "top" | "bottom";
+  className?: string;
+  isPausing?: boolean;
 }
 
 export function ServerStatusContextMenu({
@@ -21,10 +26,15 @@ export function ServerStatusContextMenu({
   onStartServer,
   conversationStatus,
   position = "top",
+  className = "",
+  isPausing = false,
 }: ServerStatusContextMenuProps) {
   const { t } = useTranslation();
   const ref = useClickOutsideElement<HTMLUListElement>(onClose);
 
+  const shouldActionShown =
+    conversationStatus === "RUNNING" || conversationStatus === "STOPPED";
+
   return (
     <ContextMenu
       ref={ref}
@@ -32,24 +42,36 @@ export function ServerStatusContextMenu({
       position={position}
       alignment="left"
       size="default"
-      className="left-2 w-fit min-w-max"
+      className={cn("left-2 w-fit min-w-42", className)}
     >
-      {conversationStatus === "RUNNING" && onStopServer && (
-        <ServerStatusContextMenuIconText
-          icon={<StopCircleIcon width={18} height={18} />}
-          text={t(I18nKey.COMMON$STOP_RUNTIME)}
-          onClick={onStopServer}
-          testId="stop-server-button"
-        />
-      )}
+      <ServerStatus
+        conversationStatus={conversationStatus}
+        isPausing={isPausing}
+        className="py-1"
+      />
 
-      {conversationStatus === "STOPPED" && onStartServer && (
-        <ServerStatusContextMenuIconText
-          icon={<PlayCircleIcon width={18} height={18} />}
-          text={t(I18nKey.COMMON$START_RUNTIME)}
-          onClick={onStartServer}
-          testId="start-server-button"
-        />
+      {shouldActionShown && (
+        <>
+          <Divider />
+
+          {conversationStatus === "RUNNING" && onStopServer && (
+            <ServerStatusContextMenuIconText
+              icon={<StopCircleIcon width={18} height={18} />}
+              text={t(I18nKey.COMMON$STOP_RUNTIME)}
+              onClick={onStopServer}
+              testId="stop-server-button"
+            />
+          )}
+
+          {conversationStatus === "STOPPED" && onStartServer && (
+            <ServerStatusContextMenuIconText
+              icon={<PlayCircleIcon width={18} height={18} />}
+              text={t(I18nKey.COMMON$START_RUNTIME)}
+              onClick={onStartServer}
+              testId="start-server-button"
+            />
+          )}
+        </>
       )}
     </ContextMenu>
   );

frontend/src/components/features/controls/server-status.tsx

@@ -1,30 +1,23 @@
 import { useTranslation } from "react-i18next";
-import { useState } from "react";
 import DebugStackframeDot from "#/icons/debug-stackframe-dot.svg?react";
 import { I18nKey } from "#/i18n/declaration";
 import { ConversationStatus } from "#/types/conversation-status";
 import { AgentState } from "#/types/agent-state";
-import { ServerStatusContextMenu } from "./server-status-context-menu";
 import { useAgentState } from "#/hooks/use-agent-state";
 import { useTaskPolling } from "#/hooks/query/use-task-polling";
+import { getStatusColor } from "#/utils/utils";
 
 export interface ServerStatusProps {
   className?: string;
   conversationStatus: ConversationStatus | null;
   isPausing?: boolean;
-  handleStop: () => void;
-  handleResumeAgent: () => void;
 }
 
 export function ServerStatus({
   className = "",
   conversationStatus,
   isPausing = false,
-  handleStop,
-  handleResumeAgent,
 }: ServerStatusProps) {
-  const [showContextMenu, setShowContextMenu] = useState(false);
-
   const { curAgentState } = useAgentState();
   const { t } = useTranslation();
   const { isTask, taskStatus, taskDetail } = useTaskPolling();
@@ -34,34 +27,15 @@ export function ServerStatus({
 
   const isStopStatus = conversationStatus === "STOPPED";
 
-  // Get the appropriate color based on agent status
-  const getStatusColor = (): string => {
-    // Show pausing status
-    if (isPausing) {
-      return "#FFD600";
-    }
+  const statusColor = getStatusColor({
+    isPausing,
+    isTask,
+    taskStatus,
+    isStartingStatus,
+    isStopStatus,
+    curAgentState,
+  });
 
-    // Show task status if we're polling a task
-    if (isTask && taskStatus) {
-      if (taskStatus === "ERROR") {
-        return "#FF684E";
-      }
-      return "#FFD600";
-    }
-
-    if (isStartingStatus) {
-      return "#FFD600";
-    }
-    if (isStopStatus) {
-      return "#ffffff";
-    }
-    if (curAgentState === AgentState.ERROR) {
-      return "#FF684E";
-    }
-    return "#BCFF8C";
-  };
-
-  // Get the appropriate status text based on agent status
   const getStatusText = (): string => {
     // Show pausing status
     if (isPausing) {
@@ -100,49 +74,14 @@ export function ServerStatus({
     return t(I18nKey.COMMON$RUNNING);
   };
 
-  const handleClick = () => {
-    if (conversationStatus === "RUNNING" || conversationStatus === "STOPPED") {
-      setShowContextMenu(true);
-    }
-  };
-
-  const handleCloseContextMenu = () => {
-    setShowContextMenu(false);
-  };
-
-  const handleStopServer = (event: React.MouseEvent<HTMLButtonElement>) => {
-    event.preventDefault();
-    handleStop();
-    setShowContextMenu(false);
-  };
-
-  const handleStartServer = (event: React.MouseEvent<HTMLButtonElement>) => {
-    event.preventDefault();
-    handleResumeAgent();
-    setShowContextMenu(false);
-  };
-
-  const statusColor = getStatusColor();
   const statusText = getStatusText();
 
   return (
-    <div className={`relative ${className}`}>
-      <div className="flex items-center cursor-pointer" onClick={handleClick}>
+    <div className={className} data-testid="server-status">
+      <div className="flex items-center">
         <DebugStackframeDot className="w-6 h-6" color={statusColor} />
-        <span className="text-[11px] text-white font-normal leading-5">
-          {statusText}
-        </span>
+        <span className="text-[13px] text-white font-normal">{statusText}</span>
       </div>
-
-      {showContextMenu && (
-        <ServerStatusContextMenu
-          onClose={handleCloseContextMenu}
-          onStopServer={handleStopServer}
-          onStartServer={handleStartServer}
-          conversationStatus={conversationStatus}
-          position="top"
-        />
-      )}
     </div>
   );
 }

frontend/src/components/features/conversation/conversation-name-with-status.tsx

@@ -0,0 +1,79 @@
+import React from "react";
+import { useParams } from "react-router";
+import { useAgentState } from "#/hooks/use-agent-state";
+import { useTaskPolling } from "#/hooks/query/use-task-polling";
+import { useActiveConversation } from "#/hooks/query/use-active-conversation";
+import { useUnifiedPauseConversationSandbox } from "#/hooks/mutation/use-unified-stop-conversation";
+import { useUnifiedResumeConversationSandbox } from "#/hooks/mutation/use-unified-start-conversation";
+import { useUserProviders } from "#/hooks/use-user-providers";
+import { getStatusColor } from "#/utils/utils";
+import { AgentState } from "#/types/agent-state";
+import DebugStackframeDot from "#/icons/debug-stackframe-dot.svg?react";
+import { ServerStatusContextMenu } from "../controls/server-status-context-menu";
+import { ConversationName } from "./conversation-name";
+
+export function ConversationNameWithStatus() {
+  const { conversationId } = useParams<{ conversationId: string }>();
+  const { data: conversation } = useActiveConversation();
+  const { curAgentState } = useAgentState();
+  const { isTask, taskStatus } = useTaskPolling();
+  const { mutate: pauseConversationSandbox } =
+    useUnifiedPauseConversationSandbox();
+  const { mutate: resumeConversationSandbox } =
+    useUnifiedResumeConversationSandbox();
+  const { providers } = useUserProviders();
+
+  const isStartingStatus =
+    curAgentState === AgentState.LOADING || curAgentState === AgentState.INIT;
+  const isStopStatus = conversation?.status === "STOPPED";
+
+  const statusColor = getStatusColor({
+    isPausing: false,
+    isTask,
+    taskStatus,
+    isStartingStatus,
+    isStopStatus,
+    curAgentState,
+  });
+
+  const handleStopServer = (event: React.MouseEvent<HTMLButtonElement>) => {
+    event.preventDefault();
+    event.stopPropagation();
+    if (conversationId) {
+      pauseConversationSandbox({ conversationId });
+    }
+  };
+
+  const handleStartServer = (event: React.MouseEvent<HTMLButtonElement>) => {
+    event.preventDefault();
+    event.stopPropagation();
+    if (conversationId) {
+      resumeConversationSandbox({ conversationId, providers });
+    }
+  };
+
+  return (
+    <div className="flex items-center">
+      <div className="group relative">
+        <DebugStackframeDot
+          className="ml-[3.5px] w-6 h-6 cursor-pointer"
+          color={statusColor}
+        />
+        <ServerStatusContextMenu
+          onClose={() => {}}
+          onStopServer={
+            conversation?.status === "RUNNING" ? handleStopServer : undefined
+          }
+          onStartServer={
+            conversation?.status === "STOPPED" ? handleStartServer : undefined
+          }
+          conversationStatus={conversation?.status ?? null}
+          position="bottom"
+          className="opacity-0 invisible pointer-events-none group-hover:opacity-100 group-hover:visible group-hover:pointer-events-auto bottom-full left-0 mt-0 min-h-fit"
+          isPausing={false}
+        />
+      </div>
+      <ConversationName />
+    </div>
+  );
+}

frontend/src/components/features/conversation/conversation-name.tsx

@@ -124,7 +124,7 @@ export function ConversationName() {
   return (
     <>
       <div
-        className="flex items-center gap-2 h-[22px] text-base font-normal text-left pl-0 lg:pl-3.5"
+        className="flex items-center gap-2 h-[22px] text-base font-normal text-left pl-0 lg:pl-1"
         data-testid="conversation-name"
       >
         {titleMode === "edit" ? (

frontend/src/components/features/conversation/conversation-tabs/conversation-tab-content/conversation-tab-content.tsx

@@ -8,26 +8,28 @@ import { TabContentArea } from "./tab-content-area";
 import { ConversationTabTitle } from "../conversation-tab-title";
 import Terminal from "#/components/features/terminal/terminal";
 import { useConversationStore } from "#/state/conversation-store";
+import { useConversationId } from "#/hooks/use-conversation-id";
 
 // Lazy load all tab components
 const EditorTab = lazy(() => import("#/routes/changes-tab"));
 const BrowserTab = lazy(() => import("#/routes/browser-tab"));
-const JupyterTab = lazy(() => import("#/routes/jupyter-tab"));
 const ServedTab = lazy(() => import("#/routes/served-tab"));
 const VSCodeTab = lazy(() => import("#/routes/vscode-tab"));
+const PlannerTab = lazy(() => import("#/routes/planner-tab"));
 
 export function ConversationTabContent() {
   const { selectedTab, shouldShownAgentLoading } = useConversationStore();
+  const { conversationId } = useConversationId();
 
   const { t } = useTranslation();
 
   // Determine which tab is active based on the current path
   const isEditorActive = selectedTab === "editor";
   const isBrowserActive = selectedTab === "browser";
-  const isJupyterActive = selectedTab === "jupyter";
   const isServedActive = selectedTab === "served";
   const isVSCodeActive = selectedTab === "vscode";
   const isTerminalActive = selectedTab === "terminal";
+  const isPlannerActive = selectedTab === "planner";
 
   // Define tab configurations
   const tabs = [
@@ -37,11 +39,6 @@ export function ConversationTabContent() {
       component: BrowserTab,
       isActive: isBrowserActive,
     },
-    {
-      key: "jupyter",
-      component: JupyterTab,
-      isActive: isJupyterActive,
-    },
     { key: "served", component: ServedTab, isActive: isServedActive },
     { key: "vscode", component: VSCodeTab, isActive: isVSCodeActive },
     {
@@ -49,6 +46,11 @@ export function ConversationTabContent() {
       component: Terminal,
       isActive: isTerminalActive,
     },
+    {
+      key: "planner",
+      component: PlannerTab,
+      isActive: isPlannerActive,
+    },
   ];
 
   const conversationTabTitle = useMemo(() => {
@@ -58,9 +60,6 @@ export function ConversationTabContent() {
     if (isBrowserActive) {
       return t(I18nKey.COMMON$BROWSER);
     }
-    if (isJupyterActive) {
-      return t(I18nKey.COMMON$JUPYTER);
-    }
     if (isServedActive) {
       return t(I18nKey.COMMON$APP);
     }
@@ -70,14 +69,17 @@ export function ConversationTabContent() {
     if (isTerminalActive) {
       return t(I18nKey.COMMON$TERMINAL);
     }
+    if (isPlannerActive) {
+      return t(I18nKey.COMMON$PLANNER);
+    }
     return "";
   }, [
     isEditorActive,
     isBrowserActive,
-    isJupyterActive,
     isServedActive,
     isVSCodeActive,
     isTerminalActive,
+    isPlannerActive,
   ]);
 
   if (shouldShownAgentLoading) {
@@ -89,7 +91,11 @@ export function ConversationTabContent() {
       <ConversationTabTitle title={conversationTabTitle} />
       <TabContentArea>
         {tabs.map(({ key, component: Component, isActive }) => (
-          <TabWrapper key={key} isActive={isActive}>
+          <TabWrapper
+            // Force Terminal tab remount to reset XTerm buffer/state when conversationId changes
+            key={key === "terminal" ? `${key}-${conversationId}` : key}
+            isActive={isActive}
+          >
             <Component />
           </TabWrapper>
         ))}

frontend/src/components/features/conversation/conversation-tabs/conversation-tab-nav.tsx

@@ -5,12 +5,16 @@ type ConversationTabNavProps = {
   icon: ComponentType<{ className: string }>;
   onClick(): void;
   isActive?: boolean;
+  label?: string;
+  className?: string;
 };
 
 export function ConversationTabNav({
   icon: Icon,
   onClick,
   isActive,
+  label,
+  className,
 }: ConversationTabNavProps) {
   return (
     <button
@@ -19,18 +23,21 @@ export function ConversationTabNav({
         onClick();
       }}
       className={cn(
-        "p-1 rounded-md cursor-pointer",
+        "flex items-center gap-2 rounded-md cursor-pointer",
+        "pl-1.5 pr-2 py-1",
         "text-[#9299AA] bg-[#0D0F11]",
         isActive && "bg-[#25272D] text-white",
         isActive
           ? "hover:text-white hover:bg-tertiary"
           : "hover:text-white hover:bg-[#0D0F11]",
-        isActive
-          ? "focus-within:text-white focus-within:bg-tertiary"
-          : "focus-within:text-[#9299AA] focus-within:bg-[#0D0F11]",
+        isActive ? "focus-within:text-white" : "focus-within:text-[#9299AA]",
+        className,
       )}
     >
-      <Icon className={cn("w-5 h-5 text-inherit")} />
+      <Icon className={cn("w-5 h-5 text-inherit flex-shrink-0")} />
+      {isActive && label && (
+        <span className="text-sm font-medium whitespace-nowrap">{label}</span>
+      )}
     </button>
   );
 }

frontend/src/components/features/conversation/conversation-tabs/conversation-tabs-context-menu.tsx

@@ -0,0 +1,116 @@
+import { useTranslation } from "react-i18next";
+import { useLocalStorage } from "@uidotdev/usehooks";
+import { ContextMenu } from "#/ui/context-menu";
+import { ContextMenuListItem } from "../../context-menu/context-menu-list-item";
+import { useClickOutsideElement } from "#/hooks/use-click-outside-element";
+import { I18nKey } from "#/i18n/declaration";
+import TerminalIcon from "#/icons/terminal.svg?react";
+import GlobeIcon from "#/icons/globe.svg?react";
+import ServerIcon from "#/icons/server.svg?react";
+import GitChanges from "#/icons/git_changes.svg?react";
+import VSCodeIcon from "#/icons/vscode.svg?react";
+import PillIcon from "#/icons/pill.svg?react";
+import PillFillIcon from "#/icons/pill-fill.svg?react";
+import { USE_PLANNING_AGENT } from "#/utils/feature-flags";
+import LessonPlanIcon from "#/icons/lesson-plan.svg?react";
+
+interface ConversationTabsContextMenuProps {
+  isOpen: boolean;
+  onClose: () => void;
+}
+
+export function ConversationTabsContextMenu({
+  isOpen,
+  onClose,
+}: ConversationTabsContextMenuProps) {
+  const ref = useClickOutsideElement<HTMLUListElement>(onClose);
+  const { t } = useTranslation();
+  const [unpinnedTabs, setUnpinnedTabs] = useLocalStorage<string[]>(
+    "conversation-unpinned-tabs",
+    [],
+  );
+
+  const shouldUsePlanningAgent = USE_PLANNING_AGENT();
+
+  const tabConfig = [
+    {
+      tab: "editor",
+      icon: GitChanges,
+      i18nKey: I18nKey.COMMON$CHANGES,
+    },
+    {
+      tab: "vscode",
+      icon: VSCodeIcon,
+      i18nKey: I18nKey.COMMON$CODE,
+    },
+    {
+      tab: "terminal",
+      icon: TerminalIcon,
+      i18nKey: I18nKey.COMMON$TERMINAL,
+    },
+    {
+      tab: "served",
+      icon: ServerIcon,
+      i18nKey: I18nKey.COMMON$APP,
+    },
+    {
+      tab: "browser",
+      icon: GlobeIcon,
+      i18nKey: I18nKey.COMMON$BROWSER,
+    },
+  ];
+
+  if (shouldUsePlanningAgent) {
+    tabConfig.unshift({
+      tab: "planner",
+      icon: LessonPlanIcon,
+      i18nKey: I18nKey.COMMON$PLANNER,
+    });
+  }
+
+  if (!isOpen) return null;
+
+  const handleTabClick = (tab: string) => {
+    const tabString = tab;
+    if (unpinnedTabs.includes(tabString)) {
+      // Tab is unpinned, pin it (remove from unpinned list)
+      setUnpinnedTabs(
+        unpinnedTabs.filter((unpinnedTab) => unpinnedTab !== tabString),
+      );
+    } else {
+      // Tab is pinned, unpin it (add to unpinned list)
+      setUnpinnedTabs([...unpinnedTabs, tabString]);
+    }
+  };
+
+  const isTabPinned = (tab: string) => !unpinnedTabs.includes(tab as string);
+
+  return (
+    <ContextMenu
+      testId="conversation-tabs-context-menu"
+      ref={ref}
+      alignment="right"
+      position="bottom"
+      className="mt-2 w-fit z-[9999]"
+    >
+      {tabConfig.map(({ tab, icon: Icon, i18nKey }) => {
+        const pinned = isTabPinned(tab);
+        return (
+          <ContextMenuListItem
+            key={tab}
+            onClick={() => handleTabClick(tab)}
+            className="flex items-center gap-2 p-2 hover:bg-[#5C5D62] rounded h-[30px]"
+          >
+            <Icon className="w-4 h-4" />
+            <span className="text-white text-sm">{t(i18nKey)}</span>
+            {pinned ? (
+              <PillFillIcon className="w-7 h-7 ml-auto flex-shrink-0 text-white -mr-[5px]" />
+            ) : (
+              <PillIcon className="w-4.5 h-4.5 ml-auto flex-shrink-0 text-white" />
+            )}
+          </ContextMenuListItem>
+        );
+      })}
+    </ContextMenu>
+  );
+}

frontend/src/components/features/conversation/conversation-tabs/conversation-tabs.tsx

@@ -1,12 +1,13 @@
-import { useEffect } from "react";
+import { useEffect, useState } from "react";
 import { useTranslation } from "react-i18next";
 import { useLocalStorage } from "@uidotdev/usehooks";
-import JupyterIcon from "#/icons/jupyter.svg?react";
 import TerminalIcon from "#/icons/terminal.svg?react";
 import GlobeIcon from "#/icons/globe.svg?react";
 import ServerIcon from "#/icons/server.svg?react";
 import GitChanges from "#/icons/git_changes.svg?react";
 import VSCodeIcon from "#/icons/vscode.svg?react";
+import ThreeDotsVerticalIcon from "#/icons/three-dots-vertical.svg?react";
+import LessonPlanIcon from "#/icons/lesson-plan.svg?react";
 import { cn } from "#/utils/utils";
 import { ConversationTabNav } from "./conversation-tab-nav";
 import { ChatActionTooltip } from "../../chat/chat-action-tooltip";
@@ -16,6 +17,8 @@ import {
   useConversationStore,
   type ConversationTab,
 } from "#/state/conversation-store";
+import { ConversationTabsContextMenu } from "./conversation-tabs-context-menu";
+import { USE_PLANNING_AGENT } from "#/utils/feature-flags";
 
 export function ConversationTabs() {
   const {
@@ -25,6 +28,8 @@ export function ConversationTabs() {
     setSelectedTab,
   } = useConversationStore();
 
+  const [isMenuOpen, setIsMenuOpen] = useState(false);
+
   // Persist selectedTab and isRightPanelShown in localStorage
   const [persistedSelectedTab, setPersistedSelectedTab] =
     useLocalStorage<ConversationTab | null>(
@@ -35,6 +40,13 @@ export function ConversationTabs() {
   const [persistedIsRightPanelShown, setPersistedIsRightPanelShown] =
     useLocalStorage<boolean>("conversation-right-panel-shown", true);
 
+  const [persistedUnpinnedTabs] = useLocalStorage<string[]>(
+    "conversation-unpinned-tabs",
+    [],
+  );
+
+  const shouldUsePlanningAgent = USE_PLANNING_AGENT();
+
   const onTabChange = (value: ConversationTab | null) => {
     setSelectedTab(value);
     // Persist the selected tab to localStorage
@@ -88,49 +100,70 @@ export function ConversationTabs() {
 
   const tabs = [
     {
+      tabValue: "editor",
       isActive: isTabActive("editor"),
       icon: GitChanges,
       onClick: () => onTabSelected("editor"),
       tooltipContent: t(I18nKey.COMMON$CHANGES),
       tooltipAriaLabel: t(I18nKey.COMMON$CHANGES),
+      label: t(I18nKey.COMMON$CHANGES),
     },
     {
+      tabValue: "vscode",
       isActive: isTabActive("vscode"),
       icon: VSCodeIcon,
       onClick: () => onTabSelected("vscode"),
       tooltipContent: <VSCodeTooltipContent />,
       tooltipAriaLabel: t(I18nKey.COMMON$CODE),
+      label: t(I18nKey.COMMON$CODE),
     },
     {
+      tabValue: "terminal",
       isActive: isTabActive("terminal"),
       icon: TerminalIcon,
       onClick: () => onTabSelected("terminal"),
       tooltipContent: t(I18nKey.COMMON$TERMINAL),
       tooltipAriaLabel: t(I18nKey.COMMON$TERMINAL),
+      label: t(I18nKey.COMMON$TERMINAL),
+      className: "pl-2",
     },
     {
-      isActive: isTabActive("jupyter"),
-      icon: JupyterIcon,
-      onClick: () => onTabSelected("jupyter"),
-      tooltipContent: t(I18nKey.COMMON$JUPYTER),
-      tooltipAriaLabel: t(I18nKey.COMMON$JUPYTER),
-    },
-    {
+      tabValue: "served",
       isActive: isTabActive("served"),
       icon: ServerIcon,
       onClick: () => onTabSelected("served"),
       tooltipContent: t(I18nKey.COMMON$APP),
       tooltipAriaLabel: t(I18nKey.COMMON$APP),
+      label: t(I18nKey.COMMON$APP),
     },
     {
+      tabValue: "browser",
       isActive: isTabActive("browser"),
       icon: GlobeIcon,
       onClick: () => onTabSelected("browser"),
       tooltipContent: t(I18nKey.COMMON$BROWSER),
       tooltipAriaLabel: t(I18nKey.COMMON$BROWSER),
+      label: t(I18nKey.COMMON$BROWSER),
     },
   ];
 
+  if (shouldUsePlanningAgent) {
+    tabs.unshift({
+      tabValue: "planner",
+      isActive: isTabActive("planner"),
+      icon: LessonPlanIcon,
+      onClick: () => onTabSelected("planner"),
+      tooltipContent: t(I18nKey.COMMON$PLANNER),
+      tooltipAriaLabel: t(I18nKey.COMMON$PLANNER),
+      label: t(I18nKey.COMMON$PLANNER),
+    });
+  }
+
+  // Filter out unpinned tabs
+  const visibleTabs = tabs.filter(
+    (tab) => !persistedUnpinnedTabs.includes(tab.tabValue),
+  );
+
   return (
     <div
       className={cn(
@@ -138,9 +171,17 @@ export function ConversationTabs() {
         "flex flex-row justify-start lg:justify-end items-center gap-4.5",
       )}
     >
-      {tabs.map(
+      {visibleTabs.map(
         (
-          { icon, onClick, isActive, tooltipContent, tooltipAriaLabel },
+          {
+            icon,
+            onClick,
+            isActive,
+            tooltipContent,
+            tooltipAriaLabel,
+            label,
+            className,
+          },
           index,
         ) => (
           <ChatActionTooltip
@@ -152,10 +193,29 @@ export function ConversationTabs() {
               icon={icon}
               onClick={onClick}
               isActive={isActive}
+              label={label}
+              className={className}
             />
           </ChatActionTooltip>
         ),
       )}
+      <div className="relative">
+        <button
+          type="button"
+          onClick={() => setIsMenuOpen(!isMenuOpen)}
+          className={cn(
+            "p-1 pl-0 rounded-md cursor-pointer",
+            "text-[#9299AA] bg-[#0D0F11]",
+          )}
+          aria-label={t(I18nKey.COMMON$MORE_OPTIONS)}
+        >
+          <ThreeDotsVerticalIcon className={cn("w-5 h-5 text-inherit")} />
+        </button>
+        <ConversationTabsContextMenu
+          isOpen={isMenuOpen}
+          onClose={() => setIsMenuOpen(false)}
+        />
+      </div>
     </div>
   );
 }

frontend/src/components/features/diff-viewer/file-diff-viewer.tsx

@@ -7,7 +7,7 @@ import { GitChangeStatus } from "#/api/open-hands.types";
 import { getLanguageFromPath } from "#/utils/get-language-from-path";
 import { cn } from "#/utils/utils";
 import ChevronUp from "#/icons/chveron-up.svg?react";
-import { useGitDiff } from "#/hooks/query/use-get-diff";
+import { useUnifiedGitDiff } from "#/hooks/query/use-unified-git-diff";
 
 interface LoadingSpinnerProps {
   className?: string;
@@ -64,7 +64,7 @@ export function FileDiffViewer({ path, type }: FileDiffViewerProps) {
     isLoading,
     isSuccess,
     isRefetching,
-  } = useGitDiff({
+  } = useUnifiedGitDiff({
     filePath,
     type,
     enabled: !isCollapsed,

frontend/src/components/features/home/repo-selection-form.tsx

@@ -131,7 +131,7 @@ export function RepositorySelectionForm({
         onBranchSelect={handleBranchSelection}
         defaultBranch={defaultBranch}
         placeholder="Select branch..."
-        className="max-w-[500px]"
+        className="max-w-full"
         disabled={!selectedRepository || isLoadingSettings}
       />
     );

frontend/src/components/features/jupyter/jupyter-cell-input.tsx

@@ -1,22 +0,0 @@
-import SyntaxHighlighter from "react-syntax-highlighter";
-import { atomOneDark } from "react-syntax-highlighter/dist/esm/styles/hljs";
-
-interface JupytrerCellInputProps {
-  code: string;
-}
-
-export function JupytrerCellInput({ code }: JupytrerCellInputProps) {
-  return (
-    <div className="rounded-lg bg-gray-800 dark:bg-gray-900 p-2 text-xs">
-      <div className="mb-1 text-gray-400">EXECUTE</div>
-      <pre
-        className="scrollbar-custom scrollbar-thumb-gray-500 hover:scrollbar-thumb-gray-400 dark:scrollbar-thumb-white/10 dark:hover:scrollbar-thumb-white/20 overflow-auto px-5"
-        style={{ padding: 0, marginBottom: 0, fontSize: "0.75rem" }}
-      >
-        <SyntaxHighlighter language="python" style={atomOneDark} wrapLongLines>
-          {code}
-        </SyntaxHighlighter>
-      </pre>
-    </div>
-  );
-}

frontend/src/components/features/jupyter/jupyter-cell-output.tsx

@@ -1,55 +0,0 @@
-import Markdown from "react-markdown";
-import SyntaxHighlighter from "react-syntax-highlighter";
-import { atomOneDark } from "react-syntax-highlighter/dist/esm/styles/hljs";
-import { useTranslation } from "react-i18next";
-import { I18nKey } from "#/i18n/declaration";
-import { JupyterLine } from "#/utils/parse-cell-content";
-import { paragraph } from "../markdown/paragraph";
-
-interface JupyterCellOutputProps {
-  lines: JupyterLine[];
-}
-
-export function JupyterCellOutput({ lines }: JupyterCellOutputProps) {
-  const { t } = useTranslation();
-  return (
-    <div className="rounded-lg bg-gray-800 dark:bg-gray-900 p-2 text-xs">
-      <div className="mb-1 text-gray-400">
-        {t(I18nKey.JUPYTER$OUTPUT_LABEL)}
-      </div>
-      <pre
-        className="scrollbar-custom scrollbar-thumb-gray-500 hover:scrollbar-thumb-gray-400 dark:scrollbar-thumb-white/10 dark:hover:scrollbar-thumb-white/20 overflow-auto px-5 max-h-[60vh] bg-gray-800"
-        style={{ padding: 0, marginBottom: 0, fontSize: "0.75rem" }}
-      >
-        {/* display the lines as plaintext or image */}
-        {lines.map((line, index) => {
-          if (line.type === "image") {
-            // Use markdown to display the image
-            const imageMarkdown = line.url
-              ? `![image](${line.url})`
-              : line.content;
-            return (
-              <div key={index}>
-                <Markdown
-                  components={{
-                    p: paragraph,
-                  }}
-                  urlTransform={(value: string) => value}
-                >
-                  {imageMarkdown}
-                </Markdown>
-              </div>
-            );
-          }
-          return (
-            <div key={index}>
-              <SyntaxHighlighter language="plaintext" style={atomOneDark}>
-                {line.content}
-              </SyntaxHighlighter>
-            </div>
-          );
-        })}
-      </pre>
-    </div>
-  );
-}

frontend/src/components/features/jupyter/jupyter-cell.tsx

@@ -1,23 +0,0 @@
-import React from "react";
-import { Cell } from "#/state/jupyter-store";
-import { JupyterLine, parseCellContent } from "#/utils/parse-cell-content";
-import { JupytrerCellInput } from "./jupyter-cell-input";
-import { JupyterCellOutput } from "./jupyter-cell-output";
-
-interface JupyterCellProps {
-  cell: Cell;
-}
-
-export function JupyterCell({ cell }: JupyterCellProps) {
-  const [lines, setLines] = React.useState<JupyterLine[]>([]);
-
-  React.useEffect(() => {
-    setLines(parseCellContent(cell.content, cell.imageUrls));
-  }, [cell.content, cell.imageUrls]);
-
-  if (cell.type === "input") {
-    return <JupytrerCellInput code={cell.content} />;
-  }
-
-  return <JupyterCellOutput lines={lines} />;
-}

frontend/src/components/features/jupyter/jupyter.tsx

@@ -1,63 +0,0 @@
-import React from "react";
-import { useTranslation } from "react-i18next";
-import { useScrollToBottom } from "#/hooks/use-scroll-to-bottom";
-import { JupyterCell } from "./jupyter-cell";
-import { ScrollToBottomButton } from "#/components/shared/buttons/scroll-to-bottom-button";
-import { RUNTIME_INACTIVE_STATES } from "#/types/agent-state";
-import { I18nKey } from "#/i18n/declaration";
-import JupyterLargeIcon from "#/icons/jupyter-large.svg?react";
-import { WaitingForRuntimeMessage } from "../chat/waiting-for-runtime-message";
-import { useAgentState } from "#/hooks/use-agent-state";
-import { useJupyterStore } from "#/state/jupyter-store";
-
-interface JupyterEditorProps {
-  maxWidth: number;
-}
-
-export function JupyterEditor({ maxWidth }: JupyterEditorProps) {
-  const { curAgentState } = useAgentState();
-
-  const cells = useJupyterStore((state) => state.cells);
-
-  const jupyterRef = React.useRef<HTMLDivElement>(null);
-
-  const { t } = useTranslation();
-
-  const isRuntimeInactive = RUNTIME_INACTIVE_STATES.includes(curAgentState);
-
-  const { hitBottom, scrollDomToBottom, onChatBodyScroll } =
-    useScrollToBottom(jupyterRef);
-
-  return (
-    <>
-      {isRuntimeInactive && <WaitingForRuntimeMessage />}
-      {!isRuntimeInactive && cells.length > 0 && (
-        <div className="flex-1 h-full flex flex-col" style={{ maxWidth }}>
-          <div
-            data-testid="jupyter-container"
-            className="flex-1 overflow-y-auto fast-smooth-scroll custom-scrollbar-always rounded-xl"
-            ref={jupyterRef}
-            onScroll={(e) => onChatBodyScroll(e.currentTarget)}
-          >
-            {cells.map((cell, index) => (
-              <JupyterCell key={index} cell={cell} />
-            ))}
-          </div>
-          {!hitBottom && (
-            <div className="sticky bottom-2 flex items-center justify-center">
-              <ScrollToBottomButton onClick={scrollDomToBottom} />
-            </div>
-          )}
-        </div>
-      )}
-      {!isRuntimeInactive && cells.length === 0 && (
-        <div className="flex flex-col items-center justify-center w-full h-full p-10 gap-4">
-          <JupyterLargeIcon width={113} height={113} color="#A1A1A1" />
-          <span className="text-[#8D95A9] text-[19px] font-normal leading-5">
-            {t(I18nKey.COMMON$JUPYTER_EMPTY_MESSAGE)}
-          </span>
-        </div>
-      )}
-    </>
-  );
-}

frontend/src/components/features/markdown/paragraph.tsx

@@ -7,5 +7,5 @@ export function paragraph({
 }: React.ClassAttributes<HTMLParagraphElement> &
   React.HTMLAttributes<HTMLParagraphElement> &
   ExtraProps) {
-  return <p className="pb-[10px] last:pb-0">{children}</p>;
+  return <p className="py-2.5 first:pt-0 last:pb-0">{children}</p>;
 }

frontend/src/components/features/payment/payment-form.tsx

@@ -1,8 +1,7 @@
 import React from "react";
-import { useTranslation, Trans } from "react-i18next";
+import { useTranslation } from "react-i18next";
 import { useCreateStripeCheckoutSession } from "#/hooks/mutation/stripe/use-create-stripe-checkout-session";
 import { useBalance } from "#/hooks/query/use-balance";
-import { useSubscriptionAccess } from "#/hooks/query/use-subscription-access";
 import { cn } from "#/utils/utils";
 import MoneyIcon from "#/icons/money.svg?react";
 import { SettingsInput } from "../settings/settings-input";
@@ -11,24 +10,13 @@ import { LoadingSpinner } from "#/components/shared/loading-spinner";
 import { amountIsValid } from "#/utils/amount-is-valid";
 import { I18nKey } from "#/i18n/declaration";
 import { PoweredByStripeTag } from "./powered-by-stripe-tag";
-import { CancelSubscriptionModal } from "./cancel-subscription-modal";
 
 export function PaymentForm() {
   const { t } = useTranslation();
   const { data: balance, isLoading } = useBalance();
-  const { data: subscriptionAccess } = useSubscriptionAccess();
   const { mutate: addBalance, isPending } = useCreateStripeCheckoutSession();
 
   const [buttonIsDisabled, setButtonIsDisabled] = React.useState(true);
-  const [showCancelModal, setShowCancelModal] = React.useState(false);
-
-  const subscriptionExpiredDate =
-    subscriptionAccess?.end_at &&
-    new Date(subscriptionAccess.end_at).toLocaleDateString("en-US", {
-      year: "numeric",
-      month: "long",
-      day: "numeric",
-    });
 
   const billingFormAction = async (formData: FormData) => {
     const amount = formData.get("top-up-input")?.toString();
@@ -94,50 +82,7 @@ export function PaymentForm() {
           {isPending && <LoadingSpinner size="small" />}
           <PoweredByStripeTag />
         </div>
-
-        {/* Cancel Subscription Button or Cancellation Message */}
-        {subscriptionAccess && (
-          <div className="flex flex-col w-[680px] gap-2 mt-4">
-            {subscriptionAccess.cancelled_at ? (
-              <div className="text-red-500 text-sm">
-                <Trans
-                  i18nKey={I18nKey.PAYMENT$SUBSCRIPTION_CANCELLED_EXPIRES}
-                  values={{ date: subscriptionExpiredDate }}
-                  components={{ date: <span className="underline" /> }}
-                />
-              </div>
-            ) : (
-              <div className="flex items-center gap-4">
-                <BrandButton
-                  testId="cancel-subscription-button"
-                  variant="ghost-danger"
-                  type="button"
-                  onClick={() => setShowCancelModal(true)}
-                >
-                  {t(I18nKey.PAYMENT$CANCEL_SUBSCRIPTION)}
-                </BrandButton>
-                <div
-                  className="text-sm text-gray-300"
-                  data-testid="next-billing-date"
-                >
-                  <Trans
-                    i18nKey={I18nKey.PAYMENT$NEXT_BILLING_DATE}
-                    values={{ date: subscriptionExpiredDate }}
-                    components={{ date: <span className="underline" /> }}
-                  />
-                </div>
-              </div>
-            )}
-          </div>
-        )}
       </div>
-
-      {/* Cancel Subscription Modal */}
-      <CancelSubscriptionModal
-        isOpen={showCancelModal}
-        onClose={() => setShowCancelModal(false)}
-        endDate={subscriptionExpiredDate}
-      />
     </form>
   );
 }

frontend/src/components/features/settings/settings-layout.tsx

@@ -11,13 +11,11 @@ interface NavigationItem {
 interface SettingsLayoutProps {
   children: React.ReactNode;
   navigationItems: NavigationItem[];
-  isSaas: boolean;
 }
 
 export function SettingsLayout({
   children,
   navigationItems,
-  isSaas,
 }: SettingsLayoutProps) {
   const [isMobileMenuOpen, setIsMobileMenuOpen] = useState(false);
 
@@ -44,7 +42,6 @@ export function SettingsLayout({
           isMobileMenuOpen={isMobileMenuOpen}
           onCloseMobileMenu={closeMobileMenu}
           navigationItems={navigationItems}
-          isSaas={isSaas}
         />
 
         {/* Main content */}

frontend/src/components/features/settings/settings-navigation.tsx

@@ -5,7 +5,6 @@ import { Typography } from "#/ui/typography";
 import { I18nKey } from "#/i18n/declaration";
 import SettingsIcon from "#/icons/settings-gear.svg?react";
 import CloseIcon from "#/icons/close.svg?react";
-import { ProPill } from "./pro-pill";
 
 interface NavigationItem {
   to: string;
@@ -17,14 +16,12 @@ interface SettingsNavigationProps {
   isMobileMenuOpen: boolean;
   onCloseMobileMenu: () => void;
   navigationItems: NavigationItem[];
-  isSaas: boolean;
 }
 
 export function SettingsNavigation({
   isMobileMenuOpen,
   onCloseMobileMenu,
   navigationItems,
-  isSaas,
 }: SettingsNavigationProps) {
   const { t } = useTranslation();
 
@@ -85,7 +82,6 @@ export function SettingsNavigation({
                 <Typography.Text className="text-[#A3A3A3] whitespace-nowrap">
                   {t(text as I18nKey)}
                 </Typography.Text>
-                {isSaas && to === "/settings" && <ProPill />}
               </div>
             </NavLink>
           ))}

frontend/src/components/features/waitlist/auth-modal.tsx

@@ -11,6 +11,7 @@ import BitbucketLogo from "#/assets/branding/bitbucket-logo.svg?react";
 import { useAuthUrl } from "#/hooks/use-auth-url";
 import { GetConfigResponse } from "#/api/option-service/option.types";
 import { Provider } from "#/types/settings";
+import { useTracking } from "#/hooks/use-tracking";
 
 interface AuthModalProps {
   githubAuthUrl: string | null;
@@ -26,6 +27,7 @@ export function AuthModal({
   providersConfigured,
 }: AuthModalProps) {
   const { t } = useTranslation();
+  const { trackLoginButtonClick } = useTracking();
 
   const gitlabAuthUrl = useAuthUrl({
     appMode: appMode || null,
@@ -47,6 +49,7 @@ export function AuthModal({
 
   const handleGitHubAuth = () => {
     if (githubAuthUrl) {
+      trackLoginButtonClick({ provider: "github" });
       // Always start the OIDC flow, let the backend handle TOS check
       window.location.href = githubAuthUrl;
     }
@@ -54,6 +57,7 @@ export function AuthModal({
 
   const handleGitLabAuth = () => {
     if (gitlabAuthUrl) {
+      trackLoginButtonClick({ provider: "gitlab" });
       // Always start the OIDC flow, let the backend handle TOS check
       window.location.href = gitlabAuthUrl;
     }
@@ -61,6 +65,7 @@ export function AuthModal({
 
   const handleBitbucketAuth = () => {
     if (bitbucketAuthUrl) {
+      trackLoginButtonClick({ provider: "bitbucket" });
       // Always start the OIDC flow, let the backend handle TOS check
       window.location.href = bitbucketAuthUrl;
     }
@@ -68,6 +73,7 @@ export function AuthModal({
 
   const handleEnterpriseSsoAuth = () => {
     if (enterpriseSsoUrl) {
+      trackLoginButtonClick({ provider: "enterprise_sso" });
       // Always start the OIDC flow, let the backend handle TOS check
       window.location.href = enterpriseSsoUrl;
     }

frontend/src/components/shared/buttons/v1-confirmation-buttons.tsx

@@ -0,0 +1,141 @@
+import { useCallback, useEffect } from "react";
+import { useTranslation } from "react-i18next";
+import { I18nKey } from "#/i18n/declaration";
+import { AgentState } from "#/types/agent-state";
+import { ActionTooltip } from "../action-tooltip";
+import { RiskAlert } from "#/components/shared/risk-alert";
+import WarningIcon from "#/icons/u-warning.svg?react";
+import { useEventMessageStore } from "#/stores/event-message-store";
+import { useEventStore } from "#/stores/use-event-store";
+import { isV1Event, isActionEvent } from "#/types/v1/type-guards";
+import { useActiveConversation } from "#/hooks/query/use-active-conversation";
+import { useAgentState } from "#/hooks/use-agent-state";
+import { useRespondToConfirmation } from "#/hooks/mutation/use-respond-to-confirmation";
+import { SecurityRisk } from "#/types/v1/core/base/common";
+
+export function V1ConfirmationButtons() {
+  const v1SubmittedEventIds = useEventMessageStore(
+    (state) => state.v1SubmittedEventIds,
+  );
+  const addV1SubmittedEventId = useEventMessageStore(
+    (state) => state.addV1SubmittedEventId,
+  );
+
+  const { t } = useTranslation();
+  const { data: conversation } = useActiveConversation();
+  const { curAgentState } = useAgentState();
+  const { mutate: respondToConfirmation } = useRespondToConfirmation();
+  const events = useEventStore((state) => state.events);
+
+  // Find the most recent V1 action awaiting confirmation
+  const awaitingAction = events
+    .filter(isV1Event)
+    .slice()
+    .reverse()
+    .find((ev) => {
+      if (ev.source !== "agent") return false;
+      // For V1, we check if the agent state is waiting for confirmation
+      return curAgentState === AgentState.AWAITING_USER_CONFIRMATION;
+    });
+
+  const handleConfirmation = useCallback(
+    (accept: boolean) => {
+      if (!awaitingAction || !conversation) {
+        return;
+      }
+
+      // Mark event as submitted to prevent duplicate submissions
+      addV1SubmittedEventId(awaitingAction.id);
+
+      // Call the V1 API endpoint
+      respondToConfirmation({
+        conversationId: conversation.conversation_id,
+        conversationUrl: conversation.url || "",
+        sessionApiKey: conversation.session_api_key,
+        accept,
+      });
+    },
+    [
+      awaitingAction,
+      conversation,
+      addV1SubmittedEventId,
+      respondToConfirmation,
+    ],
+  );
+
+  // Handle keyboard shortcuts
+  useEffect(() => {
+    if (!awaitingAction) {
+      return undefined;
+    }
+
+    const handleCancelShortcut = (event: KeyboardEvent) => {
+      if (event.shiftKey && event.metaKey && event.key === "Backspace") {
+        event.preventDefault();
+        handleConfirmation(false);
+      }
+    };
+
+    const handleContinueShortcut = (event: KeyboardEvent) => {
+      if (event.metaKey && event.key === "Enter") {
+        event.preventDefault();
+        handleConfirmation(true);
+      }
+    };
+
+    const handleKeyDown = (event: KeyboardEvent) => {
+      // Cancel: Shift+Cmd+Backspace (⇧⌘⌫)
+      handleCancelShortcut(event);
+      // Continue: Cmd+Enter (⌘↩)
+      handleContinueShortcut(event);
+    };
+
+    document.addEventListener("keydown", handleKeyDown);
+
+    return () => document.removeEventListener("keydown", handleKeyDown);
+  }, [awaitingAction, handleConfirmation]);
+
+  // Only show if agent is waiting for confirmation and we haven't already submitted
+  if (
+    curAgentState !== AgentState.AWAITING_USER_CONFIRMATION ||
+    !awaitingAction ||
+    v1SubmittedEventIds.includes(awaitingAction.id)
+  ) {
+    return null;
+  }
+
+  // Get security risk from the action (only ActionEvent has security_risk)
+  const risk = isActionEvent(awaitingAction)
+    ? awaitingAction.security_risk
+    : SecurityRisk.UNKNOWN;
+
+  const isHighRisk = risk === SecurityRisk.HIGH;
+
+  return (
+    <div className="flex flex-col gap-2 pt-4">
+      {isHighRisk && (
+        <RiskAlert
+          content={t(I18nKey.CHAT_INTERFACE$HIGH_RISK_WARNING)}
+          icon={<WarningIcon width={16} height={16} color="#fff" />}
+          severity="high"
+          title={t(I18nKey.COMMON$HIGH_RISK)}
+        />
+      )}
+      <div className="flex justify-between items-center">
+        <p className="text-sm font-normal text-white">
+          {t(I18nKey.CHAT_INTERFACE$USER_ASK_CONFIRMATION)}
+        </p>
+        <div className="flex items-center gap-3">
+          <ActionTooltip
+            type="reject"
+            onClick={() => handleConfirmation(false)}
+          />
+          <ActionTooltip
+            type="confirm"
+            onClick={() => handleConfirmation(true)}
+          />
+        </div>
+      </div>
+    </div>
+  );
+}

frontend/src/components/v1/chat/event-content-helpers/get-event-content.tsx

@@ -134,9 +134,16 @@ const getObservationEventTitle = (event: OpenHandsEvent): React.ReactNode => {
     case "BrowserObservation":
       observationKey = "OBSERVATION_MESSAGE$BROWSE";
       break;
-    case "TaskTrackerObservation":
-      observationKey = "OBSERVATION_MESSAGE$TASK_TRACKING";
+    case "TaskTrackerObservation": {
+      const { command } = event.observation;
+      if (command === "plan") {
+        observationKey = "OBSERVATION_MESSAGE$TASK_TRACKING_PLAN";
+      } else {
+        // command === "view"
+        observationKey = "OBSERVATION_MESSAGE$TASK_TRACKING_VIEW";
+      }
       break;
+    }
     default:
       // For unknown observations, use the type name
       return observationType.replace("Observation", "").toUpperCase();

frontend/src/components/v1/chat/event-content-helpers/get-observation-content.ts

@@ -19,6 +19,10 @@ const getFileEditorObservationContent = (
 ): string => {
   const { observation } = event;
 
+  if (observation.error) {
+    return `**Error:**\n${observation.error}`;
+  }
+
   const successMessage = getObservationResult(event) === "success";
 
   // For view commands or successful edits with content changes, format as code block

frontend/src/components/v1/chat/event-content-helpers/get-observation-result.ts

@@ -11,9 +11,10 @@ export const getObservationResult = (
   switch (observationType) {
     case "ExecuteBashObservation": {
       const exitCode = observation.exit_code;
+      const { metadata } = observation;
 
-      if (exitCode === -1) return "timeout"; // Command timed out
-      if (exitCode === 0) return "success"; // Command executed successfully
+      if (exitCode === -1 || metadata.exit_code === -1) return "timeout"; // Command timed out
+      if (exitCode === 0 || metadata.exit_code === 0) return "success"; // Command executed successfully
       return "error"; // Command failed
     }
     case "FileEditorObservation":

frontend/src/components/v1/chat/event-content-helpers/should-render-event.ts

@@ -7,17 +7,6 @@ import {
   isConversationStateUpdateEvent,
 } from "#/types/v1/type-guards";
 
-// V1 events that should not be rendered
-const NO_RENDER_ACTION_TYPES = [
-  "ThinkAction",
-  // Add more action types that should not be rendered
-];
-
-const NO_RENDER_OBSERVATION_TYPES = [
-  "ThinkObservation",
-  // Add more observation types that should not be rendered
-];
-
 export const shouldRenderEvent = (event: OpenHandsEvent) => {
   // Explicitly exclude system events that should not be rendered in chat
   if (isConversationStateUpdateEvent(event)) {
@@ -34,18 +23,12 @@ export const shouldRenderEvent = (event: OpenHandsEvent) => {
       return false;
     }
 
-    return !NO_RENDER_ACTION_TYPES.includes(actionType);
+    return true;
   }
 
-  // Render observation events (with filtering)
+  // Render observation events
   if (isObservationEvent(event)) {
-    // For V1, observation is an object with kind property
-    const observationType = event.observation.kind;
-
-    // Note: ObservationEvent source is always "environment", not "user"
-    // So no need to check for user source here
-
-    return !NO_RENDER_OBSERVATION_TYPES.includes(observationType);
+    return true;
   }
 
   // Render message events (user and assistant messages)

frontend/src/components/v1/chat/event-message-components/generic-event-message-wrapper.tsx

@@ -1,19 +1,18 @@
-import React from "react";
 import { OpenHandsEvent } from "#/types/v1/core";
 import { GenericEventMessage } from "../../../features/chat/generic-event-message";
 import { getEventContent } from "../event-content-helpers/get-event-content";
 import { getObservationResult } from "../event-content-helpers/get-observation-result";
 import { isObservationEvent } from "#/types/v1/type-guards";
-import { ConfirmationButtons } from "#/components/shared/buttons/confirmation-buttons";
+import { V1ConfirmationButtons } from "#/components/shared/buttons/v1-confirmation-buttons";
 
 interface GenericEventMessageWrapperProps {
   event: OpenHandsEvent;
-  shouldShowConfirmationButtons: boolean;
+  isLastMessage: boolean;
 }
 
 export function GenericEventMessageWrapper({
   event,
-  shouldShowConfirmationButtons,
+  isLastMessage,
 }: GenericEventMessageWrapperProps) {
   const { title, details } = getEventContent(event);
 
@@ -27,7 +26,7 @@ export function GenericEventMessageWrapper({
         }
         initiallyExpanded={false}
       />
-      {shouldShowConfirmationButtons && <ConfirmationButtons />}
+      {isLastMessage && <V1ConfirmationButtons />}
     </div>
   );
 }

frontend/src/components/v1/chat/event-message-components/index.ts

@@ -3,3 +3,4 @@ export { ObservationPairEventMessage } from "./observation-pair-event-message";
 export { ErrorEventMessage } from "./error-event-message";
 export { FinishEventMessage } from "./finish-event-message";
 export { GenericEventMessageWrapper } from "./generic-event-message-wrapper";
+export { ThoughtEventMessage } from "./thought-event-message";

frontend/src/components/v1/chat/event-message-components/thought-event-message.tsx

@@ -0,0 +1,32 @@
+import React from "react";
+import { ActionEvent } from "#/types/v1/core";
+import { ChatMessage } from "../../../features/chat/chat-message";
+
+interface ThoughtEventMessageProps {
+  event: ActionEvent;
+  actions?: Array<{
+    icon: React.ReactNode;
+    onClick: () => void;
+    tooltip?: string;
+  }>;
+}
+
+export function ThoughtEventMessage({
+  event,
+  actions,
+}: ThoughtEventMessageProps) {
+  // Extract thought content from the action event
+  const thoughtContent = event.thought
+    .filter((t) => t.type === "text")
+    .map((t) => t.text)
+    .join("\n");
+
+  // If there's no thought content, don't render anything
+  if (!thoughtContent) {
+    return null;
+  }
+
+  return (
+    <ChatMessage type="agent" message={thoughtContent} actions={actions} />
+  );
+}

frontend/src/components/v1/chat/event-message-components/user-assistant-event-message.tsx

@@ -4,7 +4,7 @@ import { ChatMessage } from "../../../features/chat/chat-message";
 import { ImageCarousel } from "../../../features/images/image-carousel";
 // TODO: Implement file_urls support for V1 messages
 // import { FileList } from "../../../features/files/file-list";
-import { ConfirmationButtons } from "#/components/shared/buttons/confirmation-buttons";
+import { V1ConfirmationButtons } from "#/components/shared/buttons/v1-confirmation-buttons";
 import { MicroagentStatusWrapper } from "../../../features/chat/event-message-components/microagent-status-wrapper";
 // TODO: Implement V1 LikertScaleWrapper when API supports V1 event IDs
 // import { LikertScaleWrapper } from "../../../features/chat/event-message-components/likert-scale-wrapper";
@@ -13,7 +13,6 @@ import { MicroagentStatus } from "#/types/microagent-status";
 
 interface UserAssistantEventMessageProps {
   event: MessageEvent;
-  shouldShowConfirmationButtons: boolean;
   microagentStatus?: MicroagentStatus | null;
   microagentConversationId?: string;
   microagentPRUrl?: string;
@@ -22,15 +21,16 @@ interface UserAssistantEventMessageProps {
     onClick: () => void;
     tooltip?: string;
   }>;
+  isLastMessage: boolean;
 }
 
 export function UserAssistantEventMessage({
   event,
-  shouldShowConfirmationButtons,
   microagentStatus,
   microagentConversationId,
   microagentPRUrl,
   actions,
+  isLastMessage,
 }: UserAssistantEventMessageProps) {
   const message = parseMessageFromEvent(event);
 
@@ -51,7 +51,7 @@ export function UserAssistantEventMessage({
           <ImageCarousel size="small" images={imageUrls} />
         )}
         {/* TODO: Handle file_urls if V1 messages support them */}
-        {shouldShowConfirmationButtons && <ConfirmationButtons />}
+        {isLastMessage && <V1ConfirmationButtons />}
       </ChatMessage>
       <MicroagentStatusWrapper
         microagentStatus={microagentStatus}

frontend/src/components/v1/chat/event-message.tsx

@@ -14,14 +14,13 @@ import {
   ErrorEventMessage,
   UserAssistantEventMessage,
   FinishEventMessage,
-  ObservationPairEventMessage,
   GenericEventMessageWrapper,
+  ThoughtEventMessage,
 } from "./event-message-components";
 
 interface EventMessageProps {
   event: OpenHandsEvent;
-  hasObservationPair: boolean;
-  isAwaitingUserConfirmation: boolean;
+  messages: OpenHandsEvent[];
   isLastMessage: boolean;
   microagentStatus?: MicroagentStatus | null;
   microagentConversationId?: string;
@@ -37,8 +36,7 @@ interface EventMessageProps {
 /* eslint-disable react/jsx-props-no-spreading */
 export function EventMessage({
   event,
-  hasObservationPair,
-  isAwaitingUserConfirmation,
+  messages,
   isLastMessage,
   microagentStatus,
   microagentConversationId,
@@ -46,9 +44,6 @@ export function EventMessage({
   actions,
   isInLast10Actions,
 }: EventMessageProps) {
-  const shouldShowConfirmationButtons =
-    isLastMessage && event.source === "agent" && isAwaitingUserConfirmation;
-
   const { data: config } = useConfig();
 
   // V1 events use string IDs, but useFeedbackExists expects number
@@ -74,19 +69,6 @@ export function EventMessage({
     return <ErrorEventMessage event={event} {...commonProps} />;
   }
 
-  // Observation pairs with actions
-  if (hasObservationPair && isActionEvent(event)) {
-    return (
-      <ObservationPairEventMessage
-        event={event}
-        microagentStatus={microagentStatus}
-        microagentConversationId={microagentConversationId}
-        microagentPRUrl={microagentPRUrl}
-        actions={actions}
-      />
-    );
-  }
-
   // Finish actions
   if (isActionEvent(event) && event.action.kind === "FinishAction") {
     return (
@@ -97,23 +79,53 @@ export function EventMessage({
     );
   }
 
+  // Action events - render thought + action (will be replaced by thought + observation)
+  if (isActionEvent(event)) {
+    return (
+      <>
+        <ThoughtEventMessage event={event} actions={actions} />
+        <GenericEventMessageWrapper
+          event={event}
+          isLastMessage={isLastMessage}
+        />
+      </>
+    );
+  }
+
+  // Observation events - find the corresponding action and render thought + observation
+  if (isObservationEvent(event)) {
+    // Find the action that this observation is responding to
+    const correspondingAction = messages.find(
+      (msg) => isActionEvent(msg) && msg.id === event.action_id,
+    );
+
+    return (
+      <>
+        {correspondingAction && isActionEvent(correspondingAction) && (
+          <ThoughtEventMessage event={correspondingAction} actions={actions} />
+        )}
+        <GenericEventMessageWrapper
+          event={event}
+          isLastMessage={isLastMessage}
+        />
+      </>
+    );
+  }
+
   // Message events (user and assistant messages)
   if (!isActionEvent(event) && !isObservationEvent(event)) {
     // This is a MessageEvent
     return (
       <UserAssistantEventMessage
         event={event as MessageEvent}
-        shouldShowConfirmationButtons={shouldShowConfirmationButtons}
         {...commonProps}
+        isLastMessage={isLastMessage}
       />
     );
   }
 
-  // Generic fallback for all other events (including observation events)
+  // Generic fallback for all other events
   return (
-    <GenericEventMessageWrapper
-      event={event}
-      shouldShowConfirmationButtons={shouldShowConfirmationButtons}
-    />
+    <GenericEventMessageWrapper event={event} isLastMessage={isLastMessage} />
   );
 }

frontend/src/components/v1/chat/messages.tsx

@@ -1,6 +1,5 @@
 import React from "react";
 import { OpenHandsEvent } from "#/types/v1/core";
-import { isActionEvent, isObservationEvent } from "#/types/v1/type-guards";
 import { EventMessage } from "./event-message";
 import { ChatMessage } from "../../features/chat/chat-message";
 import { useOptimisticUserMessageStore } from "#/stores/optimistic-user-message-store";
@@ -9,30 +8,16 @@ import { useOptimisticUserMessageStore } from "#/stores/optimistic-user-message-
 // import MemoryIcon from "#/icons/memory_icon.svg?react";
 
 interface MessagesProps {
-  messages: OpenHandsEvent[];
-  isAwaitingUserConfirmation: boolean;
+  messages: OpenHandsEvent[]; // UI events (actions replaced by observations)
+  allEvents: OpenHandsEvent[]; // Full event history (for action lookup)
 }
 
 export const Messages: React.FC<MessagesProps> = React.memo(
-  ({ messages, isAwaitingUserConfirmation }) => {
+  ({ messages, allEvents }) => {
     const { getOptimisticUserMessage } = useOptimisticUserMessageStore();
 
     const optimisticUserMessage = getOptimisticUserMessage();
 
-    const actionHasObservationPair = React.useCallback(
-      (event: OpenHandsEvent): boolean => {
-        if (isActionEvent(event)) {
-          // Check if there's a corresponding observation event
-          return !!messages.some(
-            (msg) => isObservationEvent(msg) && msg.action_id === event.id,
-          );
-        }
-
-        return false;
-      },
-      [messages],
-    );
-
     // TODO: Implement microagent functionality for V1 if needed
     // For now, we'll skip microagent features
 
@@ -42,8 +27,7 @@ export const Messages: React.FC<MessagesProps> = React.memo(
           <EventMessage
             key={message.id}
             event={message}
-            hasObservationPair={actionHasObservationPair(message)}
-            isAwaitingUserConfirmation={isAwaitingUserConfirmation}
+            messages={allEvents}
             isLastMessage={messages.length - 1 === index}
             isInLast10Actions={messages.length - 1 - index < 10}
             // Microagent props - not implemented yet for V1

frontend/src/contexts/conversation-websocket-context.tsx

@@ -5,6 +5,7 @@ import React, {
   useState,
   useCallback,
   useMemo,
+  useRef,
 } from "react";
 import { useQueryClient } from "@tanstack/react-query";
 import { useWebSocket, WebSocketHookOptions } from "#/hooks/use-websocket";
@@ -27,6 +28,7 @@ import {
 import { handleActionEventCacheInvalidation } from "#/utils/cache-utils";
 import { buildWebSocketUrl } from "#/utils/websocket-url";
 import type { V1SendMessageRequest } from "#/api/conversation-service/v1-conversation-service.types";
+import EventService from "#/api/event-service/event-service.api";
 
 // eslint-disable-next-line @typescript-eslint/naming-convention
 export type V1_WebSocketConnectionState =
@@ -38,6 +40,7 @@ export type V1_WebSocketConnectionState =
 interface ConversationWebSocketContextType {
   connectionState: V1_WebSocketConnectionState;
   sendMessage: (message: V1SendMessageRequest) => Promise<void>;
+  isLoadingHistory: boolean;
 }
 
 const ConversationWebSocketContext = createContext<
@@ -64,9 +67,16 @@ export function ConversationWebSocketProvider({
   const { addEvent } = useEventStore();
   const { setErrorMessage, removeErrorMessage } = useErrorMessageStore();
   const { removeOptimisticUserMessage } = useOptimisticUserMessageStore();
-  const { setAgentStatus } = useV1ConversationStateStore();
+  const { setExecutionStatus } = useV1ConversationStateStore();
   const { appendInput, appendOutput } = useCommandStore();
 
+  // History loading state
+  const [isLoadingHistory, setIsLoadingHistory] = useState(true);
+  const [expectedEventCount, setExpectedEventCount] = useState<number | null>(
+    null,
+  );
+  const receivedEventCountRef = useRef(0);
+
   // Build WebSocket URL from props
   // Only build URL if we have both conversationId and conversationUrl
   // This prevents connection attempts during task polling phase
@@ -78,16 +88,43 @@ export function ConversationWebSocketProvider({
     return buildWebSocketUrl(conversationId, conversationUrl);
   }, [conversationId, conversationUrl]);
 
-  // Reset hasConnected flag when conversation changes
+  // Reset hasConnected flag and history loading state when conversation changes
   useEffect(() => {
     hasConnectedRef.current = false;
+    setIsLoadingHistory(true);
+    setExpectedEventCount(null);
+    receivedEventCountRef.current = 0;
   }, [conversationId]);
 
+  // Check if we've received all events when expectedEventCount becomes available
+  useEffect(() => {
+    if (
+      expectedEventCount !== null &&
+      receivedEventCountRef.current >= expectedEventCount &&
+      isLoadingHistory
+    ) {
+      setIsLoadingHistory(false);
+    }
+  }, [expectedEventCount, isLoadingHistory]);
+
   const handleMessage = useCallback(
     (messageEvent: MessageEvent) => {
       try {
         const event = JSON.parse(messageEvent.data);
 
+        // Track received events for history loading (count ALL events from WebSocket)
+        // Always count when loading, even if we don't have the expected count yet
+        if (isLoadingHistory) {
+          receivedEventCountRef.current += 1;
+
+          if (
+            expectedEventCount !== null &&
+            receivedEventCountRef.current >= expectedEventCount
+          ) {
+            setIsLoadingHistory(false);
+          }
+        }
+
         // Use type guard to validate v1 event structure
         if (isV1Event(event)) {
           addEvent(event);
@@ -117,10 +154,10 @@ export function ConversationWebSocketProvider({
           // TODO: Tests
           if (isConversationStateUpdateEvent(event)) {
             if (isFullStateConversationStateUpdateEvent(event)) {
-              setAgentStatus(event.value.agent_status);
+              setExecutionStatus(event.value.execution_status);
             }
             if (isAgentStatusConversationStateUpdateEvent(event)) {
-              setAgentStatus(event.value);
+              setExecutionStatus(event.value);
             }
           }
 
@@ -141,11 +178,13 @@ export function ConversationWebSocketProvider({
     },
     [
       addEvent,
+      isLoadingHistory,
+      expectedEventCount,
       setErrorMessage,
       removeOptimisticUserMessage,
       queryClient,
       conversationId,
-      setAgentStatus,
+      setExecutionStatus,
       appendInput,
       appendOutput,
     ],
@@ -164,10 +203,26 @@ export function ConversationWebSocketProvider({
     return {
       queryParams,
       reconnect: { enabled: true },
-      onOpen: () => {
+      onOpen: async () => {
         setConnectionState("OPEN");
         hasConnectedRef.current = true; // Mark that we've successfully connected
         removeErrorMessage(); // Clear any previous error messages on successful connection
+
+        // Fetch expected event count for history loading detection
+        if (conversationId) {
+          try {
+            const count = await EventService.getEventCount(conversationId);
+            setExpectedEventCount(count);
+
+            // If no events expected, mark as loaded immediately
+            if (count === 0) {
+              setIsLoadingHistory(false);
+            }
+          } catch (error) {
+            // Fall back to marking as loaded to avoid infinite loading state
+            setIsLoadingHistory(false);
+          }
+        }
       },
       onClose: (event: CloseEvent) => {
         setConnectionState("CLOSED");
@@ -188,7 +243,13 @@ export function ConversationWebSocketProvider({
       },
       onMessage: handleMessage,
     };
-  }, [handleMessage, setErrorMessage, removeErrorMessage, sessionApiKey]);
+  }, [
+    handleMessage,
+    setErrorMessage,
+    removeErrorMessage,
+    sessionApiKey,
+    conversationId,
+  ]);
 
   // Only attempt WebSocket connection when we have a valid URL
   // This prevents connection attempts during task polling phase
@@ -246,8 +307,8 @@ export function ConversationWebSocketProvider({
   }, [socket, wsUrl]);
 
   const contextValue = useMemo(
-    () => ({ connectionState, sendMessage }),
-    [connectionState, sendMessage],
+    () => ({ connectionState, sendMessage, isLoadingHistory }),
+    [connectionState, sendMessage, isLoadingHistory],
   );
 
   return (

frontend/src/hooks/mutation/conversation-mutation-utils.ts

@@ -2,6 +2,7 @@ import { QueryClient } from "@tanstack/react-query";
 import { Provider } from "#/types/settings";
 import ConversationService from "#/api/conversation-service/conversation-service.api";
 import V1ConversationService from "#/api/conversation-service/v1-conversation-service.api";
+import { SandboxService } from "#/api/sandbox-service/sandbox-service.api";
 
 /**
  * Gets the conversation version from the cache
@@ -48,7 +49,7 @@ const fetchV1ConversationData = async (
  */
 export const pauseV1ConversationSandbox = async (conversationId: string) => {
   const { sandboxId } = await fetchV1ConversationData(conversationId);
-  return V1ConversationService.pauseSandbox(sandboxId);
+  return SandboxService.pauseSandbox(sandboxId);
 };
 
 /**
@@ -75,7 +76,7 @@ export const stopV0Conversation = async (conversationId: string) =>
  */
 export const resumeV1ConversationSandbox = async (conversationId: string) => {
   const { sandboxId } = await fetchV1ConversationData(conversationId);
-  return V1ConversationService.resumeSandbox(sandboxId);
+  return SandboxService.resumeSandbox(sandboxId);
 };
 
 /**

frontend/src/hooks/mutation/use-add-git-providers.ts

@@ -1,9 +1,11 @@
 import { useMutation, useQueryClient } from "@tanstack/react-query";
 import { SecretsService } from "#/api/secrets-service";
 import { Provider, ProviderToken } from "#/types/settings";
+import { useTracking } from "#/hooks/use-tracking";
 
 export const useAddGitProviders = () => {
   const queryClient = useQueryClient();
+  const { trackGitProviderConnected } = useTracking();
 
   return useMutation({
     mutationFn: ({
@@ -11,7 +13,18 @@ export const useAddGitProviders = () => {
     }: {
       providers: Record<Provider, ProviderToken>;
     }) => SecretsService.addGitProvider(providers),
-    onSuccess: async () => {
+    onSuccess: async (_, { providers }) => {
+      // Track which providers were connected (filter out empty tokens)
+      const connectedProviders = Object.entries(providers)
+        .filter(([, value]) => value.token && value.token.trim() !== "")
+        .map(([key]) => key);
+
+      if (connectedProviders.length > 0) {
+        trackGitProviderConnected({
+          providers: connectedProviders,
+        });
+      }
+
       await queryClient.invalidateQueries({ queryKey: ["settings"] });
     },
     meta: {

frontend/src/hooks/mutation/use-create-conversation.ts

@@ -1,11 +1,11 @@
 import { useMutation, useQueryClient } from "@tanstack/react-query";
-import posthog from "posthog-js";
 import ConversationService from "#/api/conversation-service/conversation-service.api";
 import V1ConversationService from "#/api/conversation-service/v1-conversation-service.api";
 import { SuggestedTask } from "#/utils/types";
 import { Provider } from "#/types/settings";
 import { CreateMicroagent, Conversation } from "#/api/open-hands.types";
 import { USE_V1_CONVERSATION_API } from "#/utils/feature-flags";
+import { useTracking } from "#/hooks/use-tracking";
 
 interface CreateConversationVariables {
   query?: string;
@@ -31,6 +31,7 @@ interface CreateConversationResponse extends Partial<Conversation> {
 
 export const useCreateConversation = () => {
   const queryClient = useQueryClient();
+  const { trackConversationCreated } = useTracking();
 
   return useMutation({
     mutationKey: ["create-conversation"],
@@ -86,12 +87,11 @@ export const useCreateConversation = () => {
         is_v1: false,
       };
     },
-    onSuccess: async (_, { query, repository }) => {
-      posthog.capture("initial_query_submitted", {
-        entry_point: "task_form",
-        query_character_length: query?.length,
-        has_repository: !!repository,
+    onSuccess: async (_, { repository }) => {
+      trackConversationCreated({
+        hasRepository: !!repository,
       });
+
       queryClient.removeQueries({
         queryKey: ["user", "conversations"],
       });

frontend/src/hooks/mutation/use-respond-to-confirmation.ts

@@ -0,0 +1,32 @@
+import { useMutation } from "@tanstack/react-query";
+import EventService from "#/api/event-service/event-service.api";
+import type { ConfirmationResponseRequest } from "#/api/event-service/event-service.types";
+
+interface UseRespondToConfirmationVariables {
+  conversationId: string;
+  conversationUrl: string;
+  sessionApiKey?: string | null;
+  accept: boolean;
+}
+
+export const useRespondToConfirmation = () =>
+  useMutation({
+    mutationKey: ["respond-to-confirmation"],
+    mutationFn: async ({
+      conversationId,
+      conversationUrl,
+      sessionApiKey,
+      accept,
+    }: UseRespondToConfirmationVariables) => {
+      const request: ConfirmationResponseRequest = {
+        accept,
+      };
+
+      return EventService.respondToConfirmation(
+        conversationId,
+        conversationUrl,
+        request,
+        sessionApiKey,
+      );
+    },
+  });

frontend/src/hooks/mutation/use-update-conversation.ts

@@ -26,6 +26,13 @@ export const useUpdateConversation = () => {
           ),
       );
 
+      // Also optimistically update the active conversation query
+      queryClient.setQueryData(
+        ["user", "conversation", variables.conversationId],
+        (old: { title: string } | undefined) =>
+          old ? { ...old, title: variables.newTitle } : old,
+      );
+
       return { previousConversations };
     },
     onError: (err, variables, context) => {

frontend/src/hooks/query/use-batch-app-conversations.ts

@@ -0,0 +1,11 @@
+import { useQuery } from "@tanstack/react-query";
+import V1ConversationService from "#/api/conversation-service/v1-conversation-service.api";
+
+export const useBatchAppConversations = (ids: string[]) =>
+  useQuery({
+    queryKey: ["v1-batch-get-app-conversations", ids],
+    queryFn: () => V1ConversationService.batchGetAppConversations(ids),
+    enabled: ids.length > 0,
+    staleTime: 1000 * 60 * 5, // 5 minutes
+    gcTime: 1000 * 60 * 15, // 15 minutes
+  });

frontend/src/hooks/query/use-batch-feedback.ts

@@ -4,6 +4,7 @@ import ConversationService from "#/api/conversation-service/conversation-service
 import { useConversationId } from "#/hooks/use-conversation-id";
 import { useConfig } from "#/hooks/query/use-config";
 import { useRuntimeIsReady } from "#/hooks/use-runtime-is-ready";
+import { useActiveConversation } from "#/hooks/query/use-active-conversation";
 
 export interface BatchFeedbackData {
   exists: boolean;
@@ -25,13 +26,20 @@ export const getFeedbackExistsQueryKey = (
 export const useBatchFeedback = () => {
   const { conversationId } = useConversationId();
   const { data: config } = useConfig();
+  const { data: conversation } = useActiveConversation();
   const queryClient = useQueryClient();
   const runtimeIsReady = useRuntimeIsReady();
 
+  const isV1Conversation = conversation?.conversation_version === "V1";
+
   const query = useQuery({
     queryKey: getFeedbackQueryKey(conversationId),
     queryFn: () => ConversationService.getBatchFeedback(conversationId!),
-    enabled: runtimeIsReady && !!conversationId && config?.APP_MODE === "saas",
+    enabled:
+      runtimeIsReady &&
+      !!conversationId &&
+      config?.APP_MODE === "saas" &&
+      !isV1Conversation,
     staleTime: 1000 * 60 * 5, // 5 minutes
     gcTime: 1000 * 60 * 15, // 15 minutes
   });

frontend/src/hooks/query/use-batch-sandboxes.ts

@@ -0,0 +1,11 @@
+import { useQuery } from "@tanstack/react-query";
+import { SandboxService } from "#/api/sandbox-service/sandbox-service.api";
+
+export const useBatchSandboxes = (ids: string[]) =>
+  useQuery({
+    queryKey: ["sandboxes", "batch", ids],
+    queryFn: () => SandboxService.batchGetSandboxes(ids),
+    enabled: ids.length > 0,
+    staleTime: 1000 * 60 * 5, // 5 minutes
+    gcTime: 1000 * 60 * 15, // 15 minutes
+  });

frontend/src/hooks/query/use-conversation-config.ts

@@ -2,14 +2,20 @@ import { useQuery } from "@tanstack/react-query";
 import React from "react";
 import { useConversationId } from "#/hooks/use-conversation-id";
 import ConversationService from "#/api/conversation-service/conversation-service.api";
+import V1ConversationService from "#/api/conversation-service/v1-conversation-service.api";
 import { useRuntimeIsReady } from "../use-runtime-is-ready";
+import { useActiveConversation } from "./use-active-conversation";
 
-export const useConversationConfig = () => {
+/**
+ * @deprecated This hook is for V0 conversations only. Use useUnifiedConversationConfig instead,
+ * or useV1ConversationConfig once we fully migrate to V1.
+ */
+export const useV0ConversationConfig = () => {
   const { conversationId } = useConversationId();
   const runtimeIsReady = useRuntimeIsReady();
 
   const query = useQuery({
-    queryKey: ["conversation_config", conversationId],
+    queryKey: ["v0_conversation_config", conversationId],
     queryFn: () => {
       if (!conversationId) throw new Error("No conversation ID");
       return ConversationService.getRuntimeId(conversationId);
@@ -34,3 +40,80 @@ export const useConversationConfig = () => {
 
   return query;
 };
+
+export const useV1ConversationConfig = () => {
+  const { conversationId } = useConversationId();
+  const runtimeIsReady = useRuntimeIsReady();
+
+  const query = useQuery({
+    queryKey: ["v1_conversation_config", conversationId],
+    queryFn: () => {
+      if (!conversationId) throw new Error("No conversation ID");
+      return V1ConversationService.getConversationConfig(conversationId);
+    },
+    enabled: runtimeIsReady && !!conversationId,
+    staleTime: 1000 * 60 * 5, // 5 minutes
+    gcTime: 1000 * 60 * 15, // 15 minutes
+  });
+
+  React.useEffect(() => {
+    if (query.data) {
+      const { runtime_id: runtimeId } = query.data;
+
+      // eslint-disable-next-line no-console
+      console.log(
+        "Runtime ID: %c%s",
+        "background: #444; color: #ffeb3b; font-weight: bold; padding: 2px 4px; border-radius: 4px;",
+        runtimeId,
+      );
+    }
+  }, [query.data]);
+
+  return query;
+};
+
+/**
+ * Unified hook that switches between V0 and V1 conversation config endpoints based on conversation version.
+ *
+ * @temporary This hook is temporary during the V0 to V1 migration period.
+ * Once we fully migrate to V1, all code should use useV1ConversationConfig directly.
+ */
+export const useUnifiedConversationConfig = () => {
+  const { conversationId } = useConversationId();
+  const { data: conversation } = useActiveConversation();
+  const runtimeIsReady = useRuntimeIsReady();
+  const isV1Conversation = conversation?.conversation_version === "V1";
+
+  const query = useQuery({
+    queryKey: ["conversation_config", conversationId, isV1Conversation],
+    queryFn: () => {
+      if (!conversationId) throw new Error("No conversation ID");
+
+      if (isV1Conversation) {
+        return V1ConversationService.getConversationConfig(conversationId);
+      }
+      return ConversationService.getRuntimeId(conversationId);
+    },
+    enabled: runtimeIsReady && !!conversationId && conversation !== undefined,
+    staleTime: 1000 * 60 * 5, // 5 minutes
+    gcTime: 1000 * 60 * 15, // 15 minutes
+  });
+
+  React.useEffect(() => {
+    if (query.data) {
+      const { runtime_id: runtimeId } = query.data;
+
+      // eslint-disable-next-line no-console
+      console.log(
+        "Runtime ID: %c%s",
+        "background: #444; color: #ffeb3b; font-weight: bold; padding: 2px 4px; border-radius: 4px;",
+        runtimeId,
+      );
+    }
+  }, [query.data]);
+
+  return query;
+};
+
+// Keep the old export name for backward compatibility (uses unified approach)
+export const useConversationConfig = useUnifiedConversationConfig;

frontend/src/hooks/query/use-feedback-exists.ts

@@ -1,6 +1,7 @@
 import { useQuery, useQueryClient } from "@tanstack/react-query";
 import { useConversationId } from "#/hooks/use-conversation-id";
 import { useConfig } from "#/hooks/query/use-config";
+import { useActiveConversation } from "#/hooks/query/use-active-conversation";
 import { BatchFeedbackData, getFeedbackQueryKey } from "./use-batch-feedback";
 
 export type FeedbackData = BatchFeedbackData;
@@ -9,6 +10,9 @@ export const useFeedbackExists = (eventId?: number) => {
   const queryClient = useQueryClient();
   const { conversationId } = useConversationId();
   const { data: config } = useConfig();
+  const { data: conversation } = useActiveConversation();
+
+  const isV1Conversation = conversation?.conversation_version === "V1";
 
   return useQuery<FeedbackData>({
     queryKey: [...getFeedbackQueryKey(conversationId), eventId],
@@ -22,7 +26,11 @@ export const useFeedbackExists = (eventId?: number) => {
 
       return batchData?.[eventId.toString()] ?? { exists: false };
     },
-    enabled: !!eventId && !!conversationId && config?.APP_MODE === "saas",
+    enabled:
+      !!eventId &&
+      !!conversationId &&
+      config?.APP_MODE === "saas" &&
+      !isV1Conversation,
     staleTime: 1000 * 60 * 5, // 5 minutes
     gcTime: 1000 * 60 * 15, // 15 minutes
   });

frontend/src/hooks/query/use-unified-active-host.ts

@@ -0,0 +1,99 @@
+import { useQueries, useQuery } from "@tanstack/react-query";
+import axios from "axios";
+import React from "react";
+import ConversationService from "#/api/conversation-service/conversation-service.api";
+import { useConversationId } from "#/hooks/use-conversation-id";
+import { useRuntimeIsReady } from "#/hooks/use-runtime-is-ready";
+import { useActiveConversation } from "#/hooks/query/use-active-conversation";
+import { useBatchSandboxes } from "./use-batch-sandboxes";
+import { useConversationConfig } from "./use-conversation-config";
+
+/**
+ * Unified hook to get active web host for both legacy (V0) and V1 conversations
+ * - V0: Uses the legacy getWebHosts API endpoint and polls them
+ * - V1: Gets worker URLs from sandbox exposed_urls (WORKER_1, WORKER_2, etc.)
+ */
+export const useUnifiedActiveHost = () => {
+  const [activeHost, setActiveHost] = React.useState<string | null>(null);
+  const { conversationId } = useConversationId();
+  const runtimeIsReady = useRuntimeIsReady();
+  const { data: conversation } = useActiveConversation();
+  const { data: conversationConfig, isLoading: isLoadingConfig } =
+    useConversationConfig();
+
+  const isV1Conversation = conversation?.conversation_version === "V1";
+  const sandboxId = conversationConfig?.runtime_id;
+
+  // Fetch sandbox data for V1 conversations
+  const sandboxesQuery = useBatchSandboxes(sandboxId ? [sandboxId] : []);
+
+  // Get worker URLs from V1 sandbox or legacy web hosts from V0
+  const { data, isLoading: hostsQueryLoading } = useQuery({
+    queryKey: [conversationId, "unified", "hosts", isV1Conversation, sandboxId],
+    queryFn: async () => {
+      // V1: Get worker URLs from sandbox exposed_urls
+      if (isV1Conversation) {
+        if (
+          !sandboxesQuery.data ||
+          sandboxesQuery.data.length === 0 ||
+          !sandboxesQuery.data[0]
+        ) {
+          return { hosts: [] };
+        }
+
+        const sandbox = sandboxesQuery.data[0];
+        const workerUrls =
+          sandbox.exposed_urls
+            ?.filter((url) => url.name.startsWith("WORKER_"))
+            .map((url) => url.url) || [];
+
+        return { hosts: workerUrls };
+      }
+
+      // V0 (Legacy): Use the legacy API endpoint
+      const hosts = await ConversationService.getWebHosts(conversationId);
+      return { hosts };
+    },
+    enabled:
+      runtimeIsReady &&
+      !!conversationId &&
+      (!isV1Conversation || !!sandboxesQuery.data),
+    initialData: { hosts: [] },
+    meta: {
+      disableToast: true,
+    },
+  });
+
+  // Poll all hosts to find which one is active
+  const apps = useQueries({
+    queries: data.hosts.map((host) => ({
+      queryKey: [conversationId, "unified", "hosts", host],
+      queryFn: async () => {
+        try {
+          await axios.get(host);
+          return host;
+        } catch (e) {
+          return "";
+        }
+      },
+      refetchInterval: 3000,
+      meta: {
+        disableToast: true,
+      },
+    })),
+  });
+
+  const appsData = apps.map((app) => app.data);
+
+  React.useEffect(() => {
+    const successfulApp = appsData.find((app) => app);
+    setActiveHost(successfulApp || "");
+  }, [appsData]);
+
+  // Calculate overall loading state including dependent queries for V1
+  const isLoading = isV1Conversation
+    ? isLoadingConfig || sandboxesQuery.isLoading || hostsQueryLoading
+    : hostsQueryLoading;
+
+  return { activeHost, isLoading };
+};

frontend/src/hooks/query/use-unified-get-git-changes.ts

@@ -0,0 +1,107 @@
+import React from "react";
+import { useQuery } from "@tanstack/react-query";
+import GitService from "#/api/git-service/git-service.api";
+import V1GitService from "#/api/git-service/v1-git-service.api";
+import { useConversationId } from "#/hooks/use-conversation-id";
+import { useActiveConversation } from "#/hooks/query/use-active-conversation";
+import { useRuntimeIsReady } from "#/hooks/use-runtime-is-ready";
+import { getGitPath } from "#/utils/get-git-path";
+import type { GitChange } from "#/api/open-hands.types";
+
+/**
+ * Unified hook to get git changes for both legacy (V0) and V1 conversations
+ * - V0: Uses the legacy GitService.getGitChanges API endpoint
+ * - V1: Uses the V1GitService.getGitChanges API endpoint with runtime URL
+ */
+export const useUnifiedGetGitChanges = () => {
+  const { conversationId } = useConversationId();
+  const { data: conversation } = useActiveConversation();
+  const [orderedChanges, setOrderedChanges] = React.useState<GitChange[]>([]);
+  const previousDataRef = React.useRef<GitChange[] | null>(null);
+  const runtimeIsReady = useRuntimeIsReady();
+
+  const isV1Conversation = conversation?.conversation_version === "V1";
+  const conversationUrl = conversation?.url;
+  const sessionApiKey = conversation?.session_api_key;
+  const selectedRepository = conversation?.selected_repository;
+
+  // Calculate git path based on selected repository
+  const gitPath = React.useMemo(
+    () => getGitPath(selectedRepository),
+    [selectedRepository],
+  );
+
+  const result = useQuery({
+    queryKey: [
+      "file_changes",
+      conversationId,
+      isV1Conversation,
+      conversationUrl,
+      gitPath,
+    ],
+    queryFn: async () => {
+      if (!conversationId) throw new Error("No conversation ID");
+
+      // V1: Use the V1 API endpoint with runtime URL
+      if (isV1Conversation) {
+        return V1GitService.getGitChanges(
+          conversationUrl,
+          sessionApiKey,
+          gitPath,
+        );
+      }
+
+      // V0 (Legacy): Use the legacy API endpoint
+      return GitService.getGitChanges(conversationId);
+    },
+    retry: false,
+    staleTime: 1000 * 60 * 5, // 5 minutes
+    gcTime: 1000 * 60 * 15, // 15 minutes
+    enabled: runtimeIsReady && !!conversationId,
+    meta: {
+      disableToast: true,
+    },
+  });
+
+  // Latest changes should be on top
+  React.useEffect(() => {
+    if (!result.isFetching && result.isSuccess && result.data) {
+      const currentData = result.data;
+
+      // If this is new data (not the same reference as before)
+      if (currentData !== previousDataRef.current) {
+        previousDataRef.current = currentData;
+
+        // Figure out new items by comparing with what we already have
+        if (Array.isArray(currentData)) {
+          const currentIds = new Set(currentData.map((item) => item.path));
+          const existingIds = new Set(orderedChanges.map((item) => item.path));
+
+          // Filter out items that already exist in orderedChanges
+          const newItems = currentData.filter(
+            (item) => !existingIds.has(item.path),
+          );
+
+          // Filter out items that no longer exist in the API response
+          const existingItems = orderedChanges.filter((item) =>
+            currentIds.has(item.path),
+          );
+
+          // Add new items to the beginning
+          setOrderedChanges([...newItems, ...existingItems]);
+        } else {
+          // If not an array, just use the data directly
+          setOrderedChanges([currentData]);
+        }
+      }
+    }
+  }, [result.isFetching, result.isSuccess, result.data]);
+
+  return {
+    data: orderedChanges,
+    isLoading: result.isLoading,
+    isSuccess: result.isSuccess,
+    isError: result.isError,
+    error: result.error,
+  };
+};

frontend/src/hooks/query/use-unified-git-diff.ts

@@ -0,0 +1,67 @@
+import React from "react";
+import { useQuery } from "@tanstack/react-query";
+import GitService from "#/api/git-service/git-service.api";
+import V1GitService from "#/api/git-service/v1-git-service.api";
+import { useConversationId } from "#/hooks/use-conversation-id";
+import { useActiveConversation } from "#/hooks/query/use-active-conversation";
+import { getGitPath } from "#/utils/get-git-path";
+import type { GitChangeStatus } from "#/api/open-hands.types";
+
+type UseUnifiedGitDiffConfig = {
+  filePath: string;
+  type: GitChangeStatus;
+  enabled: boolean;
+};
+
+/**
+ * Unified hook to get git diff for both legacy (V0) and V1 conversations
+ * - V0: Uses the legacy GitService.getGitChangeDiff API endpoint
+ * - V1: Uses the V1GitService.getGitChangeDiff API endpoint with runtime URL
+ */
+export const useUnifiedGitDiff = (config: UseUnifiedGitDiffConfig) => {
+  const { conversationId } = useConversationId();
+  const { data: conversation } = useActiveConversation();
+
+  const isV1Conversation = conversation?.conversation_version === "V1";
+  const conversationUrl = conversation?.url;
+  const sessionApiKey = conversation?.session_api_key;
+  const selectedRepository = conversation?.selected_repository;
+
+  // For V1, we need to convert the relative file path to an absolute path
+  // The diff endpoint expects: /workspace/project/RepoName/relative/path
+  const absoluteFilePath = React.useMemo(() => {
+    if (!isV1Conversation) return config.filePath;
+
+    const gitPath = getGitPath(selectedRepository);
+    return `${gitPath}/${config.filePath}`;
+  }, [isV1Conversation, selectedRepository, config.filePath]);
+
+  return useQuery({
+    queryKey: [
+      "file_diff",
+      conversationId,
+      config.filePath,
+      config.type,
+      isV1Conversation,
+      conversationUrl,
+    ],
+    queryFn: async () => {
+      if (!conversationId) throw new Error("No conversation ID");
+
+      // V1: Use the V1 API endpoint with runtime URL and absolute path
+      if (isV1Conversation) {
+        return V1GitService.getGitChangeDiff(
+          conversationUrl,
+          sessionApiKey,
+          absoluteFilePath,
+        );
+      }
+
+      // V0 (Legacy): Use the legacy API endpoint with relative path
+      return GitService.getGitChangeDiff(conversationId, config.filePath);
+    },
+    enabled: config.enabled,
+    staleTime: 1000 * 60 * 5, // 5 minutes
+    gcTime: 1000 * 60 * 15, // 15 minutes
+  });
+};

frontend/src/hooks/query/use-unified-vscode-url.ts

@@ -0,0 +1,122 @@
+import { useQuery } from "@tanstack/react-query";
+import { useTranslation } from "react-i18next";
+import ConversationService from "#/api/conversation-service/conversation-service.api";
+import { useConversationId } from "#/hooks/use-conversation-id";
+import { useActiveConversation } from "#/hooks/query/use-active-conversation";
+import { I18nKey } from "#/i18n/declaration";
+import { transformVSCodeUrl } from "#/utils/vscode-url-helper";
+import { useRuntimeIsReady } from "#/hooks/use-runtime-is-ready";
+import { useBatchAppConversations } from "./use-batch-app-conversations";
+import { useBatchSandboxes } from "./use-batch-sandboxes";
+
+interface VSCodeUrlResult {
+  url: string | null;
+  error: string | null;
+}
+
+/**
+ * Unified hook to get VSCode URL for both legacy (V0) and V1 conversations
+ * - V0: Uses the legacy getVSCodeUrl API endpoint
+ * - V1: Gets the VSCode URL from sandbox exposed_urls
+ */
+export const useUnifiedVSCodeUrl = () => {
+  const { t } = useTranslation();
+  const { conversationId } = useConversationId();
+  const { data: conversation } = useActiveConversation();
+  const runtimeIsReady = useRuntimeIsReady();
+
+  const isV1Conversation = conversation?.conversation_version === "V1";
+
+  // Fetch V1 app conversation to get sandbox_id
+  const appConversationsQuery = useBatchAppConversations(
+    isV1Conversation && conversationId ? [conversationId] : [],
+  );
+  const appConversation = appConversationsQuery.data?.[0];
+  const sandboxId = appConversation?.sandbox_id;
+
+  // Fetch sandbox data for V1 conversations
+  const sandboxesQuery = useBatchSandboxes(sandboxId ? [sandboxId] : []);
+
+  const mainQuery = useQuery<VSCodeUrlResult>({
+    queryKey: [
+      "unified",
+      "vscode_url",
+      conversationId,
+      isV1Conversation,
+      sandboxId,
+    ],
+    queryFn: async () => {
+      if (!conversationId) throw new Error("No conversation ID");
+
+      // V1: Get VSCode URL from sandbox exposed_urls
+      if (isV1Conversation) {
+        if (
+          !sandboxesQuery.data ||
+          sandboxesQuery.data.length === 0 ||
+          !sandboxesQuery.data[0]
+        ) {
+          return {
+            url: null,
+            error: t(I18nKey.VSCODE$URL_NOT_AVAILABLE),
+          };
+        }
+
+        const sandbox = sandboxesQuery.data[0];
+        const vscodeUrl = sandbox.exposed_urls?.find(
+          (url) => url.name === "VSCODE",
+        );
+
+        if (!vscodeUrl) {
+          return {
+            url: null,
+            error: t(I18nKey.VSCODE$URL_NOT_AVAILABLE),
+          };
+        }
+
+        return {
+          url: transformVSCodeUrl(vscodeUrl.url),
+          error: null,
+        };
+      }
+
+      // V0 (Legacy): Use the legacy API endpoint
+      const data = await ConversationService.getVSCodeUrl(conversationId);
+
+      if (data.vscode_url) {
+        return {
+          url: transformVSCodeUrl(data.vscode_url),
+          error: null,
+        };
+      }
+
+      return {
+        url: null,
+        error: t(I18nKey.VSCODE$URL_NOT_AVAILABLE),
+      };
+    },
+    enabled:
+      runtimeIsReady &&
+      !!conversationId &&
+      (!isV1Conversation || !!sandboxesQuery.data),
+    refetchOnMount: true,
+    retry: 3,
+  });
+
+  // Calculate overall loading state including dependent queries for V1
+  const isLoading = isV1Conversation
+    ? appConversationsQuery.isLoading ||
+      sandboxesQuery.isLoading ||
+      mainQuery.isLoading
+    : mainQuery.isLoading;
+
+  // Explicitly destructure to avoid excessive re-renders from spreading the entire query object
+  return {
+    data: mainQuery.data,
+    error: mainQuery.error,
+    isLoading,
+    isError: mainQuery.isError,
+    isSuccess: mainQuery.isSuccess,
+    status: mainQuery.status,
+    refetch: mainQuery.refetch,
+  };
+};

frontend/src/hooks/use-agent-state.ts

@@ -3,30 +3,30 @@ import { useAgentStore } from "#/stores/agent-store";
 import { useV1ConversationStateStore } from "#/stores/v1-conversation-state-store";
 import { useActiveConversation } from "#/hooks/query/use-active-conversation";
 import { AgentState } from "#/types/agent-state";
-import { V1AgentStatus } from "#/types/v1/core/base/common";
+import { V1ExecutionStatus } from "#/types/v1/core/base/common";
 
 /**
  * Maps V1 agent status to V0 AgentState
  */
-function mapV1StatusToV0State(status: V1AgentStatus | null): AgentState {
+function mapV1StatusToV0State(status: V1ExecutionStatus | null): AgentState {
   if (!status) {
     return AgentState.LOADING;
   }
 
   switch (status) {
-    case V1AgentStatus.IDLE:
+    case V1ExecutionStatus.IDLE:
       return AgentState.AWAITING_USER_INPUT;
-    case V1AgentStatus.RUNNING:
+    case V1ExecutionStatus.RUNNING:
       return AgentState.RUNNING;
-    case V1AgentStatus.PAUSED:
+    case V1ExecutionStatus.PAUSED:
       return AgentState.PAUSED;
-    case V1AgentStatus.WAITING_FOR_CONFIRMATION:
+    case V1ExecutionStatus.WAITING_FOR_CONFIRMATION:
       return AgentState.AWAITING_USER_CONFIRMATION;
-    case V1AgentStatus.FINISHED:
+    case V1ExecutionStatus.FINISHED:
       return AgentState.FINISHED;
-    case V1AgentStatus.ERROR:
+    case V1ExecutionStatus.ERROR:
       return AgentState.ERROR;
-    case V1AgentStatus.STUCK:
+    case V1ExecutionStatus.STUCK:
       return AgentState.ERROR; // Map STUCK to ERROR for now
     default:
       return AgentState.LOADING;
@@ -41,7 +41,9 @@ function mapV1StatusToV0State(status: V1AgentStatus | null): AgentState {
 export function useAgentState() {
   const { data: conversation } = useActiveConversation();
   const v0State = useAgentStore((state) => state.curAgentState);
-  const v1Status = useV1ConversationStateStore((state) => state.agent_status);
+  const v1Status = useV1ConversationStateStore(
+    (state) => state.execution_status,
+  );
 
   const isV1Conversation = conversation?.conversation_version === "V1";