Files
AutoGPT/autogpt_platform/backend/backend/api/external/v1/tools.py
Zamil Majdy 3e25488b2d feat(copilot): add session-level dry_run flag to autopilot sessions (#12582)
## Summary
- Adds a session-level `dry_run` flag that forces ALL tool calls
(`run_block`, `run_agent`) in a copilot/autopilot session to use dry-run
simulation mode
- Stores the flag in a typed `ChatSessionMetadata` JSON model on the
`ChatSession` DB row, accessed via `session.dry_run` property
- Adds `dry_run` to the AutoPilot block Input schema so graph builders
can create dry-run autopilot nodes
- Refactors multiple copilot tools from `**kwargs` to explicit
parameters for type safety

## Changes
- **Prisma schema**: Added `metadata` JSON column to `ChatSession` model
with migration
- **Python models**: Added `ChatSessionMetadata` model with `dry_run`
field, added `metadata` field to `ChatSessionInfo` and `ChatSession`,
updated `from_db()`, `new()`, and `create_chat_session()`
- **Session propagation**: `set_execution_context(user_id, session)`
called from `baseline/service.py` so tool handlers can read
session-level flags via `session.dry_run`
- **Tool enforcement**: `run_block` and `run_agent` check
`session.dry_run` and force `dry_run=True` when set; `run_agent` blocks
scheduling in dry-run sessions
- **AutoPilot block**: Added `dry_run` input field, passes it when
creating sessions
- **Chat API**: Added `CreateSessionRequest` model with `dry_run` field
to `POST /sessions` endpoint; added `metadata` to session responses
- **Frontend**: Updated `useChatSession.ts` to pass body to the create
session mutation
- **Tool refactoring**: Multiple copilot tools refactored from
`**kwargs` to explicit named parameters (agent_browser, manage_folders,
workspace_files, connect_integration, agent_output, bash_exec, etc.) for
better type safety

## Test plan
- [x] Unit tests for `ChatSession.new()` with dry_run parameter
- [x] Unit tests for `RunBlockTool` session dry_run override
- [x] Unit tests for `RunAgentTool` session dry_run override
- [x] Unit tests for session dry_run blocks scheduling
- [x] Existing dry_run tests still pass (12/12)
- [x] Existing permissions tests still pass
- [x] All pre-commit hooks pass (ruff, isort, pyright, tsc)
- [ ] Manual: Create autopilot session with `dry_run=True`, verify
run_block/run_agent calls use simulation

---------

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-03-31 16:27:36 +00:00

153 lines
5.1 KiB
Python

"""External API routes for chat tools - stateless HTTP endpoints.
Note: These endpoints use ephemeral sessions that are not persisted to Redis.
As a result, session-based rate limiting (max_agent_runs, max_agent_schedules)
is not enforced for external API calls. Each request creates a fresh session
with zeroed counters. Rate limiting for external API consumers should be
handled separately (e.g., via API key quotas).
"""
import logging
from typing import Any
from fastapi import APIRouter, Security
from prisma.enums import APIKeyPermission
from pydantic import BaseModel, Field
from backend.api.external.middleware import require_permission
from backend.copilot.model import ChatSession
from backend.copilot.tools import find_agent_tool, run_agent_tool
from backend.copilot.tools.models import ToolResponseBase
from backend.data.auth.base import APIAuthorizationInfo
logger = logging.getLogger(__name__)
tools_router = APIRouter(prefix="/tools", tags=["tools"])
# Note: We use Security() as a function parameter dependency (auth: APIAuthorizationInfo = Security(...))
# rather than in the decorator's dependencies= list. This avoids duplicate permission checks
# while still enforcing auth AND giving us access to auth for extracting user_id.
# Request models
class FindAgentRequest(BaseModel):
query: str = Field(..., description="Search query for finding agents")
class RunAgentRequest(BaseModel):
"""Request to run or schedule an agent.
The tool automatically handles the setup flow:
- First call returns available inputs so user can decide what values to use
- Returns missing credentials if user needs to configure them
- Executes when inputs are provided OR use_defaults=true
- Schedules execution if schedule_name and cron are provided
"""
username_agent_slug: str = Field(
...,
description="The marketplace agent slug (e.g., 'username/agent-name')",
)
inputs: dict[str, Any] = Field(
default_factory=dict,
description="Dictionary of input values for the agent",
)
use_defaults: bool = Field(
default=False,
description="Set to true to run with default values (user must confirm)",
)
schedule_name: str | None = Field(
None,
description="Name for scheduled execution (triggers scheduling mode)",
)
cron: str | None = Field(
None,
description="Cron expression (5 fields: minute hour day month weekday)",
)
timezone: str = Field(
default="UTC",
description="IANA timezone (e.g., 'America/New_York', 'UTC')",
)
def _create_ephemeral_session(user_id: str) -> ChatSession:
"""Create an ephemeral session for stateless API requests."""
return ChatSession.new(user_id, dry_run=False)
@tools_router.post(
path="/find-agent",
)
async def find_agent(
request: FindAgentRequest,
auth: APIAuthorizationInfo = Security(
require_permission(APIKeyPermission.USE_TOOLS)
),
) -> dict[str, Any]:
"""
Search for agents in the marketplace based on capabilities and user needs.
Args:
request: Search query for finding agents
Returns:
List of matching agents or no results response
"""
session = _create_ephemeral_session(auth.user_id)
result = await find_agent_tool._execute(
user_id=auth.user_id,
session=session,
query=request.query,
)
return _response_to_dict(result)
@tools_router.post(
path="/run-agent",
)
async def run_agent(
request: RunAgentRequest,
auth: APIAuthorizationInfo = Security(
require_permission(APIKeyPermission.USE_TOOLS)
),
) -> dict[str, Any]:
"""
Run or schedule an agent from the marketplace.
The endpoint automatically handles the setup flow:
- Returns missing inputs if required fields are not provided
- Returns missing credentials if user needs to configure them
- Executes immediately if all requirements are met
- Schedules execution if schedule_name and cron are provided
For scheduled execution:
- Cron format: "minute hour day month weekday"
- Examples: "0 9 * * 1-5" (9am weekdays), "0 0 * * *" (daily at midnight)
- Timezone: Use IANA timezone names like "America/New_York"
Args:
request: Agent slug, inputs, and optional schedule config
Returns:
- setup_requirements: If inputs or credentials are missing
- execution_started: If agent was run or scheduled successfully
- error: If something went wrong
"""
session = _create_ephemeral_session(auth.user_id)
result = await run_agent_tool._execute(
user_id=auth.user_id,
session=session,
username_agent_slug=request.username_agent_slug,
inputs=request.inputs,
use_defaults=request.use_defaults,
schedule_name=request.schedule_name or "",
cron=request.cron or "",
timezone=request.timezone,
)
return _response_to_dict(result)
def _response_to_dict(result: ToolResponseBase) -> dict[str, Any]:
"""Convert a tool response to a dictionary for JSON serialization."""
return result.model_dump()