fix(backend/copilot): prevent duplicate block execution from pre-launch arg mismatch#12632
Merged
Conversation
…ch arg mismatch The pre-launch mechanism speculatively starts tool execution when an AssistantMessage arrives, then matches results to SDK MCP dispatches via FIFO queue. A strict arg equality check discarded pre-launched results when the SDK CLI normalised args (e.g. injecting schema defaults), causing duplicate execution of blocks with side effects like LinearCreateIssueBlock and GithubCreatePullRequestBlock. Trust FIFO ordering instead — the SDK dispatches MCP tool calls in the same order as ToolUseBlocks in the AssistantMessage. Log at debug level when args differ for observability.
github-project-automation
Bot
moved this to 🆕 Needs initial review
in AutoGPT development kanban
majdyz
commented
Mar 31, 2026
Contributor
|
Note Reviews pausedIt looks like this branch is under active development. To avoid overwhelming you with review comments due to an influx of new commits, CodeRabbit has automatically paused this review. You can configure this behavior by changing the Use the following commands to manage reviews:
Use the checkboxes below for quick actions:
WalkthroughRemoved speculative per-session tool pre-launch/queueing; tool handler now always invokes synchronous execution path. Added read-only annotations ( Changes
Sequence Diagram(s)sequenceDiagram
participant Client
participant Service
participant ToolAdapter as CopilotAdapter
participant Executor as ToolExecutor
Client->>Service: send assistant message / tool request
Service->>ToolAdapter: route to tool handler
ToolAdapter->>ToolAdapter: determine tool and annotations (readOnlyHint)
ToolAdapter->>Executor: call _execute_tool_sync(...)
Executor-->>ToolAdapter: return result / raise exception
ToolAdapter-->>Service: return MCP result or _mcp_error
Service-->>Client: deliver response
Estimated code review effort🎯 4 (Complex) | ⏱️ ~45 minutes Possibly related PRs
Suggested reviewers
Poem
🚥 Pre-merge checks | ✅ 2 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
majdyz
commented
Mar 31, 2026
Contributor
Author
|
🤖 🔵 Nit: The |
Address self-review: remove the dead arg comparison and simplify the queue from (task, args) tuples to plain tasks. The handler trusts FIFO ordering unconditionally — no args are stored or compared. Update docstring and type alias accordingly.
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
🧹 Nitpick comments (1)
autogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py (1)
522-544: Make this test prove which args actually executed.With a constant mock result, this still passes if the handler skips the queued task and performs one direct call with
"normalised"args. Make the return value depend onblock_id, or assertmock_tool.execute.await_args.kwargs["block_id"] == "original", so the test really locks in the pre-launched path.💡 One way to pin the pre-launched call
- mock_tool = _make_mock_tool("run_block", output="pre-launched-result") + async def execute_with_block_id(*_args, **kwargs): + block_id = kwargs["block_id"] + return StreamToolOutputAvailable( + toolCallId="test-id", + output=f"result-for-{block_id}", + toolName="run_block", + success=True, + ) + + mock_tool = _make_mock_tool("run_block") + mock_tool.execute = AsyncMock(side_effect=execute_with_block_id) @@ - assert "pre-launched-result" in result["content"][0]["text"] + assert "result-for-original" in result["content"][0]["text"] + assert mock_tool.execute.await_args.kwargs["block_id"] == "original"🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@autogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py` around lines 522 - 544, Update the test_arg_mismatch_uses_prelaunched_result test so it proves the pre-launched args were actually used: make the mock_tool's return depend on the incoming block_id (so "original" returns "pre-launched-result" and "normalised" would return something else) or add an assertion that inspects the executed call on mock_tool (e.g., check mock_tool.execute.await_args.kwargs["block_id"] == "original") after running handler; keep references to the existing helpers (pre_launch_tool_call and create_tool_handler) and ensure the final assertions still verify only one execution occurred (mock_tool.execute.await_count == 1) and the observed result matches the "original" block_id outcome.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Nitpick comments:
In `@autogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py`:
- Around line 522-544: Update the test_arg_mismatch_uses_prelaunched_result test
so it proves the pre-launched args were actually used: make the mock_tool's
return depend on the incoming block_id (so "original" returns
"pre-launched-result" and "normalised" would return something else) or add an
assertion that inspects the executed call on mock_tool (e.g., check
mock_tool.execute.await_args.kwargs["block_id"] == "original") after running
handler; keep references to the existing helpers (pre_launch_tool_call and
create_tool_handler) and ensure the final assertions still verify only one
execution occurred (mock_tool.execute.await_count == 1) and the observed result
matches the "original" block_id outcome.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: 58649ac0-1d79-41f1-8bcd-c95ab83b513b
📒 Files selected for processing (2)
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
📜 Review details
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (10)
- GitHub Check: check API types
- GitHub Check: Seer Code Review
- GitHub Check: type-check (3.13)
- GitHub Check: test (3.11)
- GitHub Check: test (3.13)
- GitHub Check: test (3.12)
- GitHub Check: conflicts
- GitHub Check: end-to-end tests
- GitHub Check: Analyze (python)
- GitHub Check: Check PR Status
🧰 Additional context used
📓 Path-based instructions (4)
autogpt_platform/backend/**/*.py
📄 CodeRabbit inference engine (.github/copilot-instructions.md)
autogpt_platform/backend/**/*.py: Use Python 3.11 (required; managed by Poetry via pyproject.toml) for backend development
Always run 'poetry run format' (Black + isort) before linting in backend development
Always run 'poetry run lint' (ruff) after formatting in backend developmentRefer to
@backend/CLAUDE.mdfor backend-specific commands, architecture, and development tasks
autogpt_platform/backend/**/*.py: Import only at the top level; no local/inner imports except for lazy imports of heavy optional dependencies likeopenpyxl
Use absolute imports withfrom backend.module import ...for cross-package imports; single-dot relative imports (from .sibling import ...) are acceptable for sibling modules within the same package; avoid double-dot relative imports (from ..parent import ...)
Do not use duck typing withhasattr(),getattr(), orisinstance()for type dispatch; use typed interfaces, unions, or protocols instead
Use Pydantic models for structured data instead of dataclasses, namedtuples, or dicts
Do not use linter suppressors; no# type: ignore,# noqa, or# pyright: ignorecomments — fix the underlying type/code issue instead
Use list comprehensions instead of manual loop-and-append patterns
Use early return guard clauses to avoid deep nesting
Use%sfor deferred interpolation indebuglog statements; use f-strings for readability in other log levels (e.g.,logger.debug("Processing %s items", count),logger.info(f"Processing {count} items"))
Sanitize error paths usingos.path.basename()in error messages to avoid leaking directory structure
Avoid TOCTOU (time-of-check-time-of-use) patterns; do not use check-then-act patterns for file access and credit charging operations
Use Redis pipelines withtransaction=Truefor atomicity on multi-step Redis operations
Usemax(0, value)guards for computed values that should never be negative
Keep files under ~300 lines; if a file grows beyond this, split by responsibility (extract h...
Files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
autogpt_platform/**/*.py
📄 CodeRabbit inference engine (AGENTS.md)
Format Python code with
poetry run format
Files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
autogpt_platform/backend/**/*test*.py
📄 CodeRabbit inference engine (AGENTS.md)
Run
poetry run testfor backend testing (runs pytest with docker based postgres + prisma)
Files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
autogpt_platform/backend/**/*_test.py
📄 CodeRabbit inference engine (autogpt_platform/backend/CLAUDE.md)
autogpt_platform/backend/**/*_test.py: Use pytest with snapshot testing for API responses; test files should be colocated with source files using the*_test.pynaming pattern
Mock at boundaries by mocking where the symbol is used, not where it is defined. After refactoring, update mock targets to match new module paths
UseAsyncMockfromunittest.mockfor mocking async functions
Files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
🧠 Learnings (8)
📓 Common learnings
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12356
File: autogpt_platform/backend/backend/copilot/constants.py:9-12
Timestamp: 2026-03-10T08:39:22.025Z
Learning: In Significant-Gravitas/AutoGPT PR `#12356`, the `COPILOT_SYNTHETIC_ID_PREFIX = "copilot-"` check in `create_auto_approval_record` (human_review.py) is intentional and safe. The `graph_exec_id` passed to this function comes from server-side `PendingHumanReview` DB records (not from user input); the API only accepts `node_exec_id` from users. Synthetic `copilot-*` IDs are only ever created server-side in `run_block.py`. The prefix skip avoids a DB lookup for a `AgentGraphExecution` record that legitimately does not exist for CoPilot sessions, while `user_id` scoping is enforced at the auth layer and on the resulting auto-approval record.
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12439
File: autogpt_platform/backend/backend/blocks/autogpt_copilot.py:0-0
Timestamp: 2026-03-16T17:00:02.827Z
Learning: In autogpt_platform/backend/backend/blocks/autogpt_copilot.py, the recursion guard uses two module-level ContextVars: `_copilot_recursion_depth` (tracks current nesting depth) and `_copilot_recursion_limit` (stores the chain-wide ceiling). On the first invocation, `_copilot_recursion_limit` is set to `max_recursion_depth`; nested calls use `min(inherited_limit, max_recursion_depth)`, so they can only lower the cap, never raise it. The entry/exit logic is extracted into module-level helper functions. This is the approved pattern for preventing runaway sub-agent recursion in AutogptCopilotBlock (PR `#12439`, commits 348e9f8e2 and 3b70f61b1).
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12440
File: autogpt_platform/backend/backend/copilot/workflow_import/converter.py:0-0
Timestamp: 2026-03-17T10:57:12.953Z
Learning: In Significant-Gravitas/AutoGPT PR `#12440`, `autogpt_platform/backend/backend/copilot/workflow_import/converter.py` was fully rewritten (commit 732960e2d) to no longer make direct LLM/OpenAI API calls. The converter now builds a structured text prompt for AutoPilot/CoPilot instead. There is no `response.choices` access or any direct LLM client usage in this file. Do not flag `response.choices` access or LLM client initialization patterns as issues in this file.
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12356
File: autogpt_platform/backend/backend/copilot/tools/run_block.py:297-300
Timestamp: 2026-03-10T08:38:33.249Z
Learning: In autogpt_platform/backend/backend/copilot/tools/run_block.py, the auto-approval key for sensitive block HITL review uses graph_exec_id (copilot-session-{session_id}) + node_id (copilot-node-{block_id}). This is intentional: approving a block type within a CoPilot session auto-approves all future invocations of that same block type within the same session, mirroring how auto-approve works in normal graph execution. The user explicitly opts into this session-scoped behavior via an auto-approve toggle. Without the toggle (default), each individual invocation requires its own approval.
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12213
File: autogpt_platform/frontend/src/app/api/openapi.json:9983-9995
Timestamp: 2026-02-27T15:59:00.370Z
Learning: Repo: Significant-Gravitas/AutoGPT PR: 12213 — OpenAPI/codegen
Learning: Ensuring a field is required in generated TS types needs two sides: (1) no default value on the Pydantic field, and (2) the OpenAPI model's "required" array must list it. For MCPToolInfo, making input_schema required in OpenAPI and removing Field(default_factory=dict) in the backend prevents optional typing drift.
Learnt from: Pwuts
Repo: Significant-Gravitas/AutoGPT PR: 12284
File: autogpt_platform/frontend/src/app/api/openapi.json:11897-11900
Timestamp: 2026-03-04T23:58:18.476Z
Learning: Repo: Significant-Gravitas/AutoGPT — PR `#12284`
Backend/frontend OpenAPI codegen convention: In backend/api/features/store/model.py, the StoreSubmission and StoreSubmissionAdminView models define submitted_at: datetime | None, changes_summary: str | None, and instructions: str | None with no default. This is intentional to produce “required but nullable” fields in OpenAPI (properties appear in required[] and use anyOf [type, null]). This matches Prisma’s submittedAt DateTime? and changesSummary String?. Do not flag this as a required/nullable mismatch.
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12213
File: autogpt_platform/frontend/src/app/api/openapi.json:9983-9995
Timestamp: 2026-02-27T15:59:00.370Z
Learning: Repo: Significant-Gravitas/AutoGPT PR: 12213 — Backend/frontend OpenAPI codegen
Learning: For MCP schema models, required OpenAPI fields must have no defaults in Pydantic. Specifically, MCPToolInfo.input_schema must be required (no Field(default_factory=dict)) so openapi.json emits it in "required", ensuring generated TS types treat input_schema as non-optional.
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12356
File: autogpt_platform/backend/backend/copilot/tools/run_block.py:349-370
Timestamp: 2026-03-10T08:38:36.655Z
Learning: In the AutoGPT CoPilot HITL (Human-In-The-Loop) flow (`autogpt_platform/backend/backend/copilot/tools/run_block.py`), the review card presented to users sets `editable: false`, meaning reviewers cannot modify the input payload. Therefore, credentials resolved before `is_block_exec_need_review()` remain valid and do not need to be recomputed after the review step — the original `input_data` is unchanged through the review lifecycle.
📚 Learning: 2026-03-17T10:57:12.953Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12440
File: autogpt_platform/backend/backend/copilot/workflow_import/converter.py:0-0
Timestamp: 2026-03-17T10:57:12.953Z
Learning: In Significant-Gravitas/AutoGPT PR `#12440`, `autogpt_platform/backend/backend/copilot/workflow_import/converter.py` was fully rewritten (commit 732960e2d) to no longer make direct LLM/OpenAI API calls. The converter now builds a structured text prompt for AutoPilot/CoPilot instead. There is no `response.choices` access or any direct LLM client usage in this file. Do not flag `response.choices` access or LLM client initialization patterns as issues in this file.
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
📚 Learning: 2026-02-26T17:02:22.448Z
Learnt from: Pwuts
Repo: Significant-Gravitas/AutoGPT PR: 12211
File: .pre-commit-config.yaml:160-179
Timestamp: 2026-02-26T17:02:22.448Z
Learning: Keep the pre-commit hook pattern broad for autogpt_platform/backend to ensure OpenAPI schema changes are captured. Do not narrow to backend/api/ alone, since the generated schema depends on Pydantic models across multiple directories (backend/data/, backend/blocks/, backend/copilot/, backend/integrations/, backend/util/). Narrowing could miss schema changes and cause frontend type desynchronization.
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
📚 Learning: 2026-03-04T08:04:35.881Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12273
File: autogpt_platform/backend/backend/copilot/tools/workspace_files.py:216-220
Timestamp: 2026-03-04T08:04:35.881Z
Learning: In the AutoGPT Copilot backend, ensure that SVG images are not treated as vision image types by excluding 'image/svg+xml' from INLINEABLE_MIME_TYPES and MULTIMODAL_TYPES in tool_adapter.py; the Claude API supports PNG, JPEG, GIF, and WebP for vision. SVGs (XML text) should be handled via the text path instead, not the vision path.
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
📚 Learning: 2026-03-05T15:42:08.207Z
Learnt from: ntindle
Repo: Significant-Gravitas/AutoGPT PR: 12297
File: .claude/skills/backend-check/SKILL.md:14-16
Timestamp: 2026-03-05T15:42:08.207Z
Learning: In Python files under autogpt_platform/backend (recursively), rely on poetry run format to perform formatting (Black + isort) and linting (ruff). Do not run poetry run lint as a separate step after poetry run format, since format already includes linting checks.
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
📚 Learning: 2026-03-16T16:35:40.236Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12440
File: autogpt_platform/backend/backend/api/features/workflow_import.py:54-63
Timestamp: 2026-03-16T16:35:40.236Z
Learning: Avoid using the word 'competitor' in public-facing identifiers and text. Use neutral naming for API paths, model names, function names, and UI text. Examples: rename 'CompetitorFormat' to 'SourcePlatform', 'convert_competitor_workflow' to 'convert_workflow', '/competitor-workflow' to '/workflow'. Apply this guideline to files under autogpt_platform/backend and autogpt_platform/frontend.
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
📚 Learning: 2026-03-31T15:37:38.626Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12623
File: autogpt_platform/backend/backend/copilot/tools/agent_generator/fixer.py:37-47
Timestamp: 2026-03-31T15:37:38.626Z
Learning: When validating/constructing Anthropic API model IDs in Significant-Gravitas/AutoGPT, allow the hyphen-separated Claude Opus 4.6 model ID `claude-opus-4-6` (it corresponds to `LlmModel.CLAUDE_4_6_OPUS` in `autogpt_platform/backend/backend/blocks/llm.py`). Do NOT require the dot-separated form in Anthropic contexts. Only OpenRouter routing variants should use the dot separator (e.g., `anthropic/claude-opus-4.6`); `claude-opus-4-6` should be treated as correct when passed to Anthropic, and flagged only if it’s used in the OpenRouter path where the dot form is expected.
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
📚 Learning: 2026-03-25T06:59:27.340Z
Learnt from: CR
Repo: Significant-Gravitas/AutoGPT PR: 0
File: autogpt_platform/backend/CLAUDE.md:0-0
Timestamp: 2026-03-25T06:59:27.340Z
Learning: Applies to autogpt_platform/backend/**/*_test.py : Mock at boundaries by mocking where the symbol is used, not where it is defined. After refactoring, update mock targets to match new module paths
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
🔇 Additional comments (1)
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.py (1)
467-488: Drop the debug-only arg deep-compare.Now that mismatches no longer change control flow,
launch_args != argsonly keeps the old cost of retaining and deep-comparing the whole payload. Logging that a queued task was consumed is enough here, and it would let_TaskQueueItemstop carrying largerun_blockarg dicts around.
…annotations Remove the speculative pre-launch mechanism that caused duplicate block execution (e.g. duplicate Linear tickets, GitHub PRs) due to: - Arg mismatch between AssistantMessage and SDK MCP dispatch - Broken FIFO assumption when security hooks deny a tool - Race condition where task.cancel() arrives after API call completes Instead, use the SDK's native parallel dispatch: annotate read-only tools with ToolAnnotations(readOnlyHint=True) so the CLI dispatches them concurrently. Side-effect tools (run_block, bash_exec, etc.) run sequentially — correct and safe. Removed: - pre_launch_tool_call(), cancel_pending_tool_tasks() - _tool_task_queues ContextVar and all queue logic - Pre-launch calls in service.py streaming loop - All pre-launch tests (replaced with annotation tests) Added: - _READ_ONLY_TOOLS / _READ_ONLY_E2B_TOOLS sets - ToolAnnotations(readOnlyHint=True) on 15+ read-only tools - Tests for annotation classification
majdyz
commented
Mar 31, 2026
Contributor
Author
|
🤖 🔵 Nit (Round 2): The |
Contributor
Author
|
🤖 🔵 Nit (Round 3): The |
…y property Address review: instead of a hardcoded _READ_ONLY_TOOLS frozenset, add a read_only property to BaseTool (default False) and override it to True on 15 read-only tool classes. create_copilot_mcp_server now derives annotations from base_tool.read_only so new tools only need to set the property, not edit a separate list.
Contributor
There was a problem hiding this comment.
Actionable comments posted: 2
🧹 Nitpick comments (1)
autogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py (1)
358-376: Add regression assertions for write-capable tools in read-only classification tests.Given the new readOnlyHint path, this suite should also assert that write-capable tools (notably
read_workspace_filewithsave_to_path, andbrowser_screenshot) are not treated as read-only.Possible test extension
def test_side_effect_tools_are_not_read_only(self): """Key side-effect tools should have read_only=False.""" from backend.copilot.tools import TOOL_REGISTRY - for name in ["run_block", "bash_exec", "create_agent", "run_agent"]: + for name in [ + "run_block", + "bash_exec", + "create_agent", + "run_agent", + "read_workspace_file", + "browser_screenshot", + ]: if name in TOOL_REGISTRY: assert not TOOL_REGISTRY[ name ].read_only, f"{name} should not be read_only"🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@autogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py` around lines 358 - 376, Add assertions to the read-only classification test to ensure write-capable tools are not misclassified: update test_read_only_e2b_tools_classification (or the surrounding test block referencing _READ_ONLY_E2B_TOOLS and TOOL_REGISTRY) to assert that "read_workspace_file" (when used with save_to_path capability) and "browser_screenshot" are NOT in _READ_ONLY_E2B_TOOLS and, if present in TOOL_REGISTRY, that their ToolAnnotations/readOnlyHint or tool.read_only is False; use the existing symbols _READ_ONLY_E2B_TOOLS, TOOL_REGISTRY, ToolAnnotations and the tool names "read_workspace_file" and "browser_screenshot" to locate and add these negative assertions.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@autogpt_platform/backend/backend/copilot/tools/agent_browser.py`:
- Around line 768-770: The BrowserScreenshotTool is marked read_only via the
read_only property returning True while it performs a side-effect (writes a file
using WriteWorkspaceFileTool); update the read_only property on the
BrowserScreenshotTool class to return False (or remove the read_only override so
it inherits a non-read-only default) so the tool is not classified as read-only,
and verify any callers/registrations relying on read-only behavior are adjusted
accordingly; locate the read_only property in BrowserScreenshotTool and change
its return value and/or documentation to reflect that it mutates workspace state
via WriteWorkspaceFileTool.
In `@autogpt_platform/backend/backend/copilot/tools/workspace_files.py`:
- Around line 566-568: The ReadWorkspaceFileTool is incorrectly marked read-only
which allows its mutating method _save_to_path (invoked via save_to_path) to run
in parallel; change the read_only property on the ReadWorkspaceFileTool class so
it returns False (i.e., clear that this tool is not read-only) to prevent
parallel execution of its mutating operation and ensure callers treat it as a
writer rather than a read-only task.
---
Nitpick comments:
In `@autogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py`:
- Around line 358-376: Add assertions to the read-only classification test to
ensure write-capable tools are not misclassified: update
test_read_only_e2b_tools_classification (or the surrounding test block
referencing _READ_ONLY_E2B_TOOLS and TOOL_REGISTRY) to assert that
"read_workspace_file" (when used with save_to_path capability) and
"browser_screenshot" are NOT in _READ_ONLY_E2B_TOOLS and, if present in
TOOL_REGISTRY, that their ToolAnnotations/readOnlyHint or tool.read_only is
False; use the existing symbols _READ_ONLY_E2B_TOOLS, TOOL_REGISTRY,
ToolAnnotations and the tool names "read_workspace_file" and
"browser_screenshot" to locate and add these negative assertions.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: 247a39ef-eb95-486b-9603-01d0bac1af58
📒 Files selected for processing (17)
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.pyautogpt_platform/backend/backend/copilot/tools/agent_browser.pyautogpt_platform/backend/backend/copilot/tools/agent_output.pyautogpt_platform/backend/backend/copilot/tools/base.pyautogpt_platform/backend/backend/copilot/tools/feature_requests.pyautogpt_platform/backend/backend/copilot/tools/find_agent.pyautogpt_platform/backend/backend/copilot/tools/find_block.pyautogpt_platform/backend/backend/copilot/tools/find_library_agent.pyautogpt_platform/backend/backend/copilot/tools/get_agent_building_guide.pyautogpt_platform/backend/backend/copilot/tools/get_doc_page.pyautogpt_platform/backend/backend/copilot/tools/get_mcp_guide.pyautogpt_platform/backend/backend/copilot/tools/manage_folders.pyautogpt_platform/backend/backend/copilot/tools/search_docs.pyautogpt_platform/backend/backend/copilot/tools/validate_agent.pyautogpt_platform/backend/backend/copilot/tools/web_fetch.pyautogpt_platform/backend/backend/copilot/tools/workspace_files.py
✅ Files skipped from review due to trivial changes (2)
- autogpt_platform/backend/backend/copilot/tools/validate_agent.py
- autogpt_platform/backend/backend/copilot/tools/web_fetch.py
🚧 Files skipped from review as they are similar to previous changes (1)
- autogpt_platform/backend/backend/copilot/sdk/tool_adapter.py
📜 Review details
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (11)
- GitHub Check: check API types
- GitHub Check: end-to-end tests
- GitHub Check: Analyze (typescript)
- GitHub Check: Analyze (python)
- GitHub Check: test (3.13)
- GitHub Check: test (3.11)
- GitHub Check: type-check (3.13)
- GitHub Check: test (3.12)
- GitHub Check: type-check (3.11)
- GitHub Check: Check PR Status
- GitHub Check: Seer Code Review
🧰 Additional context used
📓 Path-based instructions (4)
autogpt_platform/backend/**/*.py
📄 CodeRabbit inference engine (.github/copilot-instructions.md)
autogpt_platform/backend/**/*.py: Use Python 3.11 (required; managed by Poetry via pyproject.toml) for backend development
Always run 'poetry run format' (Black + isort) before linting in backend development
Always run 'poetry run lint' (ruff) after formatting in backend developmentRefer to
@backend/CLAUDE.mdfor backend-specific commands, architecture, and development tasks
autogpt_platform/backend/**/*.py: Import only at the top level; no local/inner imports except for lazy imports of heavy optional dependencies likeopenpyxl
Use absolute imports withfrom backend.module import ...for cross-package imports; single-dot relative imports (from .sibling import ...) are acceptable for sibling modules within the same package; avoid double-dot relative imports (from ..parent import ...)
Do not use duck typing withhasattr(),getattr(), orisinstance()for type dispatch; use typed interfaces, unions, or protocols instead
Use Pydantic models for structured data instead of dataclasses, namedtuples, or dicts
Do not use linter suppressors; no# type: ignore,# noqa, or# pyright: ignorecomments — fix the underlying type/code issue instead
Use list comprehensions instead of manual loop-and-append patterns
Use early return guard clauses to avoid deep nesting
Use%sfor deferred interpolation indebuglog statements; use f-strings for readability in other log levels (e.g.,logger.debug("Processing %s items", count),logger.info(f"Processing {count} items"))
Sanitize error paths usingos.path.basename()in error messages to avoid leaking directory structure
Avoid TOCTOU (time-of-check-time-of-use) patterns; do not use check-then-act patterns for file access and credit charging operations
Use Redis pipelines withtransaction=Truefor atomicity on multi-step Redis operations
Usemax(0, value)guards for computed values that should never be negative
Keep files under ~300 lines; if a file grows beyond this, split by responsibility (extract h...
Files:
autogpt_platform/backend/backend/copilot/tools/manage_folders.pyautogpt_platform/backend/backend/copilot/tools/agent_output.pyautogpt_platform/backend/backend/copilot/tools/find_library_agent.pyautogpt_platform/backend/backend/copilot/tools/feature_requests.pyautogpt_platform/backend/backend/copilot/tools/search_docs.pyautogpt_platform/backend/backend/copilot/tools/get_doc_page.pyautogpt_platform/backend/backend/copilot/tools/find_block.pyautogpt_platform/backend/backend/copilot/tools/get_mcp_guide.pyautogpt_platform/backend/backend/copilot/tools/agent_browser.pyautogpt_platform/backend/backend/copilot/tools/get_agent_building_guide.pyautogpt_platform/backend/backend/copilot/tools/find_agent.pyautogpt_platform/backend/backend/copilot/tools/workspace_files.pyautogpt_platform/backend/backend/copilot/tools/base.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
autogpt_platform/**/*.py
📄 CodeRabbit inference engine (AGENTS.md)
Format Python code with
poetry run format
Files:
autogpt_platform/backend/backend/copilot/tools/manage_folders.pyautogpt_platform/backend/backend/copilot/tools/agent_output.pyautogpt_platform/backend/backend/copilot/tools/find_library_agent.pyautogpt_platform/backend/backend/copilot/tools/feature_requests.pyautogpt_platform/backend/backend/copilot/tools/search_docs.pyautogpt_platform/backend/backend/copilot/tools/get_doc_page.pyautogpt_platform/backend/backend/copilot/tools/find_block.pyautogpt_platform/backend/backend/copilot/tools/get_mcp_guide.pyautogpt_platform/backend/backend/copilot/tools/agent_browser.pyautogpt_platform/backend/backend/copilot/tools/get_agent_building_guide.pyautogpt_platform/backend/backend/copilot/tools/find_agent.pyautogpt_platform/backend/backend/copilot/tools/workspace_files.pyautogpt_platform/backend/backend/copilot/tools/base.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
autogpt_platform/backend/**/*test*.py
📄 CodeRabbit inference engine (AGENTS.md)
Run
poetry run testfor backend testing (runs pytest with docker based postgres + prisma)
Files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
autogpt_platform/backend/**/*_test.py
📄 CodeRabbit inference engine (autogpt_platform/backend/CLAUDE.md)
autogpt_platform/backend/**/*_test.py: Use pytest with snapshot testing for API responses; test files should be colocated with source files using the*_test.pynaming pattern
Mock at boundaries by mocking where the symbol is used, not where it is defined. After refactoring, update mock targets to match new module paths
UseAsyncMockfromunittest.mockfor mocking async functions
Files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
🧠 Learnings (15)
📓 Common learnings
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12356
File: autogpt_platform/backend/backend/copilot/constants.py:9-12
Timestamp: 2026-03-10T08:39:22.025Z
Learning: In Significant-Gravitas/AutoGPT PR `#12356`, the `COPILOT_SYNTHETIC_ID_PREFIX = "copilot-"` check in `create_auto_approval_record` (human_review.py) is intentional and safe. The `graph_exec_id` passed to this function comes from server-side `PendingHumanReview` DB records (not from user input); the API only accepts `node_exec_id` from users. Synthetic `copilot-*` IDs are only ever created server-side in `run_block.py`. The prefix skip avoids a DB lookup for a `AgentGraphExecution` record that legitimately does not exist for CoPilot sessions, while `user_id` scoping is enforced at the auth layer and on the resulting auto-approval record.
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12440
File: autogpt_platform/backend/backend/copilot/workflow_import/converter.py:0-0
Timestamp: 2026-03-17T10:57:12.953Z
Learning: In Significant-Gravitas/AutoGPT PR `#12440`, `autogpt_platform/backend/backend/copilot/workflow_import/converter.py` was fully rewritten (commit 732960e2d) to no longer make direct LLM/OpenAI API calls. The converter now builds a structured text prompt for AutoPilot/CoPilot instead. There is no `response.choices` access or any direct LLM client usage in this file. Do not flag `response.choices` access or LLM client initialization patterns as issues in this file.
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12439
File: autogpt_platform/backend/backend/blocks/autogpt_copilot.py:0-0
Timestamp: 2026-03-16T17:00:02.827Z
Learning: In autogpt_platform/backend/backend/blocks/autogpt_copilot.py, the recursion guard uses two module-level ContextVars: `_copilot_recursion_depth` (tracks current nesting depth) and `_copilot_recursion_limit` (stores the chain-wide ceiling). On the first invocation, `_copilot_recursion_limit` is set to `max_recursion_depth`; nested calls use `min(inherited_limit, max_recursion_depth)`, so they can only lower the cap, never raise it. The entry/exit logic is extracted into module-level helper functions. This is the approved pattern for preventing runaway sub-agent recursion in AutogptCopilotBlock (PR `#12439`, commits 348e9f8e2 and 3b70f61b1).
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12213
File: autogpt_platform/frontend/src/app/api/openapi.json:9983-9995
Timestamp: 2026-02-27T15:59:00.370Z
Learning: Repo: Significant-Gravitas/AutoGPT PR: 12213 — OpenAPI/codegen
Learning: Ensuring a field is required in generated TS types needs two sides: (1) no default value on the Pydantic field, and (2) the OpenAPI model's "required" array must list it. For MCPToolInfo, making input_schema required in OpenAPI and removing Field(default_factory=dict) in the backend prevents optional typing drift.
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12356
File: autogpt_platform/backend/backend/copilot/tools/run_block.py:297-300
Timestamp: 2026-03-10T08:38:33.249Z
Learning: In autogpt_platform/backend/backend/copilot/tools/run_block.py, the auto-approval key for sensitive block HITL review uses graph_exec_id (copilot-session-{session_id}) + node_id (copilot-node-{block_id}). This is intentional: approving a block type within a CoPilot session auto-approves all future invocations of that same block type within the same session, mirroring how auto-approve works in normal graph execution. The user explicitly opts into this session-scoped behavior via an auto-approve toggle. Without the toggle (default), each individual invocation requires its own approval.
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12213
File: autogpt_platform/frontend/src/app/api/openapi.json:9983-9995
Timestamp: 2026-02-27T15:59:00.370Z
Learning: Repo: Significant-Gravitas/AutoGPT PR: 12213 — Backend/frontend OpenAPI codegen
Learning: For MCP schema models, required OpenAPI fields must have no defaults in Pydantic. Specifically, MCPToolInfo.input_schema must be required (no Field(default_factory=dict)) so openapi.json emits it in "required", ensuring generated TS types treat input_schema as non-optional.
Learnt from: Pwuts
Repo: Significant-Gravitas/AutoGPT PR: 12284
File: autogpt_platform/frontend/src/app/api/openapi.json:11897-11900
Timestamp: 2026-03-04T23:58:18.476Z
Learning: Repo: Significant-Gravitas/AutoGPT — PR `#12284`
Backend/frontend OpenAPI codegen convention: In backend/api/features/store/model.py, the StoreSubmission and StoreSubmissionAdminView models define submitted_at: datetime | None, changes_summary: str | None, and instructions: str | None with no default. This is intentional to produce “required but nullable” fields in OpenAPI (properties appear in required[] and use anyOf [type, null]). This matches Prisma’s submittedAt DateTime? and changesSummary String?. Do not flag this as a required/nullable mismatch.
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12356
File: autogpt_platform/backend/backend/copilot/tools/run_block.py:349-370
Timestamp: 2026-03-10T08:38:36.655Z
Learning: In the AutoGPT CoPilot HITL (Human-In-The-Loop) flow (`autogpt_platform/backend/backend/copilot/tools/run_block.py`), the review card presented to users sets `editable: false`, meaning reviewers cannot modify the input payload. Therefore, credentials resolved before `is_block_exec_need_review()` remain valid and do not need to be recomputed after the review step — the original `input_data` is unchanged through the review lifecycle.
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12426
File: autogpt_platform/backend/backend/copilot/sdk/service.py:0-0
Timestamp: 2026-03-15T16:52:15.463Z
Learning: In Significant-Gravitas/AutoGPT (copilot backend), GitHub tokens (GH_TOKEN / GITHUB_TOKEN) for the `gh` CLI are injected lazily per-command in `autogpt_platform/backend/backend/copilot/tools/bash_exec._execute_on_e2b()` by calling `integration_creds.get_integration_env_vars(user_id)`, not on the global SDK subprocess environment in `sdk/service.py`. This scopes credentials to individual E2B sandbox command invocations and prevents token leakage into tool output streams or uploaded transcripts.
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12445
File: autogpt_platform/backend/backend/copilot/sdk/service.py:1071-1072
Timestamp: 2026-03-17T06:48:26.471Z
Learning: In Significant-Gravitas/AutoGPT (autogpt_platform), the AI SDK enforces `z.strictObject({type, errorText})` on SSE `StreamError` responses, so additional fields like `retryable: bool` cannot be added to `StreamError` or serialized via `to_sse()`. Instead, retry signaling for transient Anthropic API errors is done via the `COPILOT_RETRYABLE_ERROR_PREFIX` constant prepended to persisted session messages (in `ChatMessage.content`). The frontend detects retryable errors by checking `markerType === "retryable_error"` from `parseSpecialMarkers()` — no SSE schema changes and no string matching on error text. This pattern was established in PR `#12445`, commit 64d82797b.
📚 Learning: 2026-02-26T17:02:22.448Z
Learnt from: Pwuts
Repo: Significant-Gravitas/AutoGPT PR: 12211
File: .pre-commit-config.yaml:160-179
Timestamp: 2026-02-26T17:02:22.448Z
Learning: Keep the pre-commit hook pattern broad for autogpt_platform/backend to ensure OpenAPI schema changes are captured. Do not narrow to backend/api/ alone, since the generated schema depends on Pydantic models across multiple directories (backend/data/, backend/blocks/, backend/copilot/, backend/integrations/, backend/util/). Narrowing could miss schema changes and cause frontend type desynchronization.
Applied to files:
autogpt_platform/backend/backend/copilot/tools/manage_folders.pyautogpt_platform/backend/backend/copilot/tools/agent_output.pyautogpt_platform/backend/backend/copilot/tools/find_library_agent.pyautogpt_platform/backend/backend/copilot/tools/feature_requests.pyautogpt_platform/backend/backend/copilot/tools/search_docs.pyautogpt_platform/backend/backend/copilot/tools/get_doc_page.pyautogpt_platform/backend/backend/copilot/tools/find_block.pyautogpt_platform/backend/backend/copilot/tools/get_mcp_guide.pyautogpt_platform/backend/backend/copilot/tools/agent_browser.pyautogpt_platform/backend/backend/copilot/tools/get_agent_building_guide.pyautogpt_platform/backend/backend/copilot/tools/find_agent.pyautogpt_platform/backend/backend/copilot/tools/workspace_files.pyautogpt_platform/backend/backend/copilot/tools/base.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
📚 Learning: 2026-03-04T08:04:35.881Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12273
File: autogpt_platform/backend/backend/copilot/tools/workspace_files.py:216-220
Timestamp: 2026-03-04T08:04:35.881Z
Learning: In the AutoGPT Copilot backend, ensure that SVG images are not treated as vision image types by excluding 'image/svg+xml' from INLINEABLE_MIME_TYPES and MULTIMODAL_TYPES in tool_adapter.py; the Claude API supports PNG, JPEG, GIF, and WebP for vision. SVGs (XML text) should be handled via the text path instead, not the vision path.
Applied to files:
autogpt_platform/backend/backend/copilot/tools/manage_folders.pyautogpt_platform/backend/backend/copilot/tools/agent_output.pyautogpt_platform/backend/backend/copilot/tools/find_library_agent.pyautogpt_platform/backend/backend/copilot/tools/feature_requests.pyautogpt_platform/backend/backend/copilot/tools/search_docs.pyautogpt_platform/backend/backend/copilot/tools/get_doc_page.pyautogpt_platform/backend/backend/copilot/tools/find_block.pyautogpt_platform/backend/backend/copilot/tools/get_mcp_guide.pyautogpt_platform/backend/backend/copilot/tools/agent_browser.pyautogpt_platform/backend/backend/copilot/tools/get_agent_building_guide.pyautogpt_platform/backend/backend/copilot/tools/find_agent.pyautogpt_platform/backend/backend/copilot/tools/workspace_files.pyautogpt_platform/backend/backend/copilot/tools/base.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
📚 Learning: 2026-03-04T12:19:39.243Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12279
File: autogpt_platform/backend/backend/copilot/tools/base.py:184-188
Timestamp: 2026-03-04T12:19:39.243Z
Learning: In autogpt_platform/backend/backend/copilot/tools/, ensure that anonymous users always pass user_id=None to tool execution methods. The anon_ prefix (e.g., anon_123) is used only for PostHog/analytics distinct_id and must not be used as an actual user_id. Use a simple truthiness check on user_id (e.g., if user_id: ... else: ... or a dedicated is_authenticated flag) to distinguish anonymous from authenticated users, and review all tool execution call sites within this directory to prevent accidentally forwarding an anon_ user_id to tools.
Applied to files:
autogpt_platform/backend/backend/copilot/tools/manage_folders.pyautogpt_platform/backend/backend/copilot/tools/agent_output.pyautogpt_platform/backend/backend/copilot/tools/find_library_agent.pyautogpt_platform/backend/backend/copilot/tools/feature_requests.pyautogpt_platform/backend/backend/copilot/tools/search_docs.pyautogpt_platform/backend/backend/copilot/tools/get_doc_page.pyautogpt_platform/backend/backend/copilot/tools/find_block.pyautogpt_platform/backend/backend/copilot/tools/get_mcp_guide.pyautogpt_platform/backend/backend/copilot/tools/agent_browser.pyautogpt_platform/backend/backend/copilot/tools/get_agent_building_guide.pyautogpt_platform/backend/backend/copilot/tools/find_agent.pyautogpt_platform/backend/backend/copilot/tools/workspace_files.pyautogpt_platform/backend/backend/copilot/tools/base.py
📚 Learning: 2026-03-31T14:22:26.566Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12622
File: autogpt_platform/backend/backend/copilot/tools/agent_search.py:223-236
Timestamp: 2026-03-31T14:22:26.566Z
Learning: In files under autogpt_platform/backend/backend/copilot/tools/, ensure agent graph enrichment uses the typed Pydantic model `backend.data.graph.Graph` for `AgentInfo.graph` (i.e., `Graph | None`), not `dict[str, Any]`. When enriching with graph data (e.g., `_enrich_agents_with_graph`), prefer calling `graph_db().get_graph(graph_id, version=None, user_id=user_id)` directly to retrieve the typed `Graph` object rather than routing through JSON conversions like `get_agent_as_json()` / `graph_to_json()`.
Applied to files:
autogpt_platform/backend/backend/copilot/tools/manage_folders.pyautogpt_platform/backend/backend/copilot/tools/agent_output.pyautogpt_platform/backend/backend/copilot/tools/find_library_agent.pyautogpt_platform/backend/backend/copilot/tools/feature_requests.pyautogpt_platform/backend/backend/copilot/tools/search_docs.pyautogpt_platform/backend/backend/copilot/tools/get_doc_page.pyautogpt_platform/backend/backend/copilot/tools/find_block.pyautogpt_platform/backend/backend/copilot/tools/get_mcp_guide.pyautogpt_platform/backend/backend/copilot/tools/agent_browser.pyautogpt_platform/backend/backend/copilot/tools/get_agent_building_guide.pyautogpt_platform/backend/backend/copilot/tools/find_agent.pyautogpt_platform/backend/backend/copilot/tools/workspace_files.pyautogpt_platform/backend/backend/copilot/tools/base.py
📚 Learning: 2026-03-05T15:42:08.207Z
Learnt from: ntindle
Repo: Significant-Gravitas/AutoGPT PR: 12297
File: .claude/skills/backend-check/SKILL.md:14-16
Timestamp: 2026-03-05T15:42:08.207Z
Learning: In Python files under autogpt_platform/backend (recursively), rely on poetry run format to perform formatting (Black + isort) and linting (ruff). Do not run poetry run lint as a separate step after poetry run format, since format already includes linting checks.
Applied to files:
autogpt_platform/backend/backend/copilot/tools/manage_folders.pyautogpt_platform/backend/backend/copilot/tools/agent_output.pyautogpt_platform/backend/backend/copilot/tools/find_library_agent.pyautogpt_platform/backend/backend/copilot/tools/feature_requests.pyautogpt_platform/backend/backend/copilot/tools/search_docs.pyautogpt_platform/backend/backend/copilot/tools/get_doc_page.pyautogpt_platform/backend/backend/copilot/tools/find_block.pyautogpt_platform/backend/backend/copilot/tools/get_mcp_guide.pyautogpt_platform/backend/backend/copilot/tools/agent_browser.pyautogpt_platform/backend/backend/copilot/tools/get_agent_building_guide.pyautogpt_platform/backend/backend/copilot/tools/find_agent.pyautogpt_platform/backend/backend/copilot/tools/workspace_files.pyautogpt_platform/backend/backend/copilot/tools/base.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
📚 Learning: 2026-03-16T16:35:40.236Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12440
File: autogpt_platform/backend/backend/api/features/workflow_import.py:54-63
Timestamp: 2026-03-16T16:35:40.236Z
Learning: Avoid using the word 'competitor' in public-facing identifiers and text. Use neutral naming for API paths, model names, function names, and UI text. Examples: rename 'CompetitorFormat' to 'SourcePlatform', 'convert_competitor_workflow' to 'convert_workflow', '/competitor-workflow' to '/workflow'. Apply this guideline to files under autogpt_platform/backend and autogpt_platform/frontend.
Applied to files:
autogpt_platform/backend/backend/copilot/tools/manage_folders.pyautogpt_platform/backend/backend/copilot/tools/agent_output.pyautogpt_platform/backend/backend/copilot/tools/find_library_agent.pyautogpt_platform/backend/backend/copilot/tools/feature_requests.pyautogpt_platform/backend/backend/copilot/tools/search_docs.pyautogpt_platform/backend/backend/copilot/tools/get_doc_page.pyautogpt_platform/backend/backend/copilot/tools/find_block.pyautogpt_platform/backend/backend/copilot/tools/get_mcp_guide.pyautogpt_platform/backend/backend/copilot/tools/agent_browser.pyautogpt_platform/backend/backend/copilot/tools/get_agent_building_guide.pyautogpt_platform/backend/backend/copilot/tools/find_agent.pyautogpt_platform/backend/backend/copilot/tools/workspace_files.pyautogpt_platform/backend/backend/copilot/tools/base.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
📚 Learning: 2026-03-31T15:37:38.626Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12623
File: autogpt_platform/backend/backend/copilot/tools/agent_generator/fixer.py:37-47
Timestamp: 2026-03-31T15:37:38.626Z
Learning: When validating/constructing Anthropic API model IDs in Significant-Gravitas/AutoGPT, allow the hyphen-separated Claude Opus 4.6 model ID `claude-opus-4-6` (it corresponds to `LlmModel.CLAUDE_4_6_OPUS` in `autogpt_platform/backend/backend/blocks/llm.py`). Do NOT require the dot-separated form in Anthropic contexts. Only OpenRouter routing variants should use the dot separator (e.g., `anthropic/claude-opus-4.6`); `claude-opus-4-6` should be treated as correct when passed to Anthropic, and flagged only if it’s used in the OpenRouter path where the dot form is expected.
Applied to files:
autogpt_platform/backend/backend/copilot/tools/manage_folders.pyautogpt_platform/backend/backend/copilot/tools/agent_output.pyautogpt_platform/backend/backend/copilot/tools/find_library_agent.pyautogpt_platform/backend/backend/copilot/tools/feature_requests.pyautogpt_platform/backend/backend/copilot/tools/search_docs.pyautogpt_platform/backend/backend/copilot/tools/get_doc_page.pyautogpt_platform/backend/backend/copilot/tools/find_block.pyautogpt_platform/backend/backend/copilot/tools/get_mcp_guide.pyautogpt_platform/backend/backend/copilot/tools/agent_browser.pyautogpt_platform/backend/backend/copilot/tools/get_agent_building_guide.pyautogpt_platform/backend/backend/copilot/tools/find_agent.pyautogpt_platform/backend/backend/copilot/tools/workspace_files.pyautogpt_platform/backend/backend/copilot/tools/base.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
📚 Learning: 2026-02-27T15:59:00.370Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12213
File: autogpt_platform/frontend/src/app/api/openapi.json:9983-9995
Timestamp: 2026-02-27T15:59:00.370Z
Learning: Repo: Significant-Gravitas/AutoGPT PR: 12213 — OpenAPI/codegen
Learning: Ensuring a field is required in generated TS types needs two sides: (1) no default value on the Pydantic field, and (2) the OpenAPI model's "required" array must list it. For MCPToolInfo, making input_schema required in OpenAPI and removing Field(default_factory=dict) in the backend prevents optional typing drift.
Applied to files:
autogpt_platform/backend/backend/copilot/tools/get_mcp_guide.pyautogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
📚 Learning: 2026-02-27T15:59:00.370Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12213
File: autogpt_platform/frontend/src/app/api/openapi.json:9983-9995
Timestamp: 2026-02-27T15:59:00.370Z
Learning: Repo: Significant-Gravitas/AutoGPT PR: 12213 — Backend/frontend OpenAPI codegen
Learning: For MCP schema models, required OpenAPI fields must have no defaults in Pydantic. Specifically, MCPToolInfo.input_schema must be required (no Field(default_factory=dict)) so openapi.json emits it in "required", ensuring generated TS types treat input_schema as non-optional.
Applied to files:
autogpt_platform/backend/backend/copilot/tools/get_mcp_guide.py
📚 Learning: 2026-03-17T10:57:12.953Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12440
File: autogpt_platform/backend/backend/copilot/workflow_import/converter.py:0-0
Timestamp: 2026-03-17T10:57:12.953Z
Learning: In Significant-Gravitas/AutoGPT PR `#12440`, `autogpt_platform/backend/backend/copilot/workflow_import/converter.py` was fully rewritten (commit 732960e2d) to no longer make direct LLM/OpenAI API calls. The converter now builds a structured text prompt for AutoPilot/CoPilot instead. There is no `response.choices` access or any direct LLM client usage in this file. Do not flag `response.choices` access or LLM client initialization patterns as issues in this file.
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
📚 Learning: 2026-03-25T06:59:27.340Z
Learnt from: CR
Repo: Significant-Gravitas/AutoGPT PR: 0
File: autogpt_platform/backend/CLAUDE.md:0-0
Timestamp: 2026-03-25T06:59:27.340Z
Learning: Applies to autogpt_platform/backend/**/*_test.py : Mock at boundaries by mocking where the symbol is used, not where it is defined. After refactoring, update mock targets to match new module paths
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
📚 Learning: 2026-02-04T16:49:42.490Z
Learnt from: CR
Repo: Significant-Gravitas/AutoGPT PR: 0
File: .github/copilot-instructions.md:0-0
Timestamp: 2026-02-04T16:49:42.490Z
Learning: Applies to autogpt_platform/backend/**/test/**/*.py : Use snapshot testing with '--snapshot-update' flag in backend tests when output changes; always review with 'git diff'
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
📚 Learning: 2026-03-25T06:59:27.340Z
Learnt from: CR
Repo: Significant-Gravitas/AutoGPT PR: 0
File: autogpt_platform/backend/CLAUDE.md:0-0
Timestamp: 2026-03-25T06:59:27.340Z
Learning: Applies to autogpt_platform/backend/**/*_test.py : Use `AsyncMock` from `unittest.mock` for mocking async functions
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
📚 Learning: 2026-03-19T15:10:53.815Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12483
File: autogpt_platform/backend/backend/copilot/tools/test_dry_run.py:298-303
Timestamp: 2026-03-19T15:10:53.815Z
Learning: In Python unittest.mock, the correct patch target depends on whether an import is eager (module-level) or lazy (inside a function/branch):
- **Module-level import** (`from foo.bar import baz` at top of file): patch where the name is used, e.g. `patch("mymodule.baz")`.
- **Lazy import** (`from foo.bar import baz` inside a function/branch, executed at call time): patch the source module, e.g. `patch("foo.bar.baz")`, because the fresh `from ... import` at call time will look up the (now-patched) name in the source module's dict.
This pattern appears in `autogpt_platform/backend/backend/copilot/tools/helpers.py` where `simulate_block` is lazily imported inside the `if dry_run:` block, making `patch("backend.executor.simulator.simulate_block")` the correct target in tests.
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py
🔇 Additional comments (13)
autogpt_platform/backend/backend/copilot/tools/base.py (1)
122-130: LGTM!Clean implementation of the
read_onlyproperty with a sensible default (False) and clear documentation explaining the MCPreadOnlyHintannotation and parallel dispatch behavior. The opt-in pattern ensures new tools are safe by default.autogpt_platform/backend/backend/copilot/tools/manage_folders.py (1)
198-200: LGTM!
ListFoldersToolcorrectly marked as read-only since it only performs list/read operations. The mutating tools (CreateFolderTool,UpdateFolderTool,MoveFolderTool,DeleteFolderTool,MoveAgentsToFolderTool) appropriately inherit the defaultFalse.autogpt_platform/backend/backend/copilot/tools/search_docs.py (1)
56-58: LGTM!Correctly marked as read-only — the tool only performs documentation searches with no side effects.
autogpt_platform/backend/backend/copilot/tools/get_doc_page.py (1)
44-46: LGTM!Correctly marked as read-only — the tool only reads documentation file content with no side effects.
autogpt_platform/backend/backend/copilot/tools/feature_requests.py (1)
152-154: LGTM!
SearchFeatureRequestsToolcorrectly marked as read-only since it only queries Linear for existing issues. TheCreateFeatureRequestToolappropriately inherits the defaultFalseas it performs mutations.autogpt_platform/backend/backend/copilot/tools/agent_output.py (1)
163-165: LGTM!Correctly marked as read-only — the tool only retrieves and views execution outputs without mutations. The
wait_for_executioncall is a polling/read operation, not a side effect.autogpt_platform/backend/backend/copilot/tools/get_agent_building_guide.py (1)
55-57: LGTM!Correctly marked as read-only — the tool only reads and returns the agent building guide content. The internal module-level caching is an implementation detail, not an external side effect.
autogpt_platform/backend/backend/copilot/tools/find_block.py (1)
81-83: LGTM!Correctly marked as read-only — the tool only searches for and retrieves block metadata with no side effects.
autogpt_platform/backend/backend/copilot/tools/find_agent.py (1)
36-38: Looks good:FindAgentToolread-only classification is appropriate.This aligns with the non-mutating behavior of the tool and fits the new parallel dispatch model.
autogpt_platform/backend/backend/copilot/tools/get_mcp_guide.py (1)
51-53:GetMCPGuideToolread-only flag is correct.Good fit for SDK readOnlyHint-based dispatch.
autogpt_platform/backend/backend/copilot/tools/workspace_files.py (1)
445-447:ListWorkspaceFilesToolread-only flag looks correct.This tool is query-only, so
read_only=Trueis a good fit.autogpt_platform/backend/backend/copilot/tools/find_library_agent.py (1)
39-41: Read-only classification forFindLibraryAgentToollooks good.Consistent with this tool’s query-only behavior.
autogpt_platform/backend/backend/copilot/sdk/tool_adapter_test.py (1)
279-333: Nice coverage update for direct-execution semantics.These tests clearly validate the post–pre-launch behavior and duplicate-execution guard.
10 tasks
Contributor
Author
E2E Test ReportDate: 2026-03-31 | Branch:
Key FindingsPre-launch mechanism fully removed: 0 log matches for readOnlyHint parallel dispatch confirmed: Two No duplicate execution: Each No regressions: CoPilot chat, tool calls, session persistence, and E2B sandbox all function correctly. Full report: |
…dOnlyHint Annotate all MCP tools (including side-effect tools like run_block, bash_exec) with readOnlyHint=True so the SDK CLI dispatches all concurrent tool calls in parallel. E2E verified: 3 bash_exec(sleep 3) calls completed in ~3.3s total (vs ~9s sequential), 3 find_block calls dispatched within 5ms.
Contributor
Author
E2E Test Report — Parallel Dispatch (all tools readOnlyHint=True)Test: 3x
|
![sentry[bot]](https://avatars.githubusercontent.com/in/12637?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 3
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@autogpt_platform/backend/backend/copilot/sdk/tool_adapter.py`:
- Around line 537-542: The code currently marks every E2B file tool as read-only
by always passing _READONLY_ANNOTATION when creating the decorated tool; change
this to consult the _READ_ONLY_E2B_TOOLS set so that writeable tools (e.g.,
write_file, edit_file) are not annotated as read-only. Locate the tool creation
call that uses tool(name, desc, schema,
annotations=_READONLY_ANNOTATION)(_truncating(handler, name)) and modify it to
conditionally include _READONLY_ANNOTATION only when name is present in
_READ_ONLY_E2B_TOOLS (otherwise omit the annotations argument or pass None),
keeping the use of _truncating(handler, name) the same. Ensure the change
references the same symbols: tool, _truncating, handler, name,
_READ_ONLY_E2B_TOOLS, and _READONLY_ANNOTATION so write/edit tools can run
non-parallelizable operations safely.
- Around line 83-86: The _READ_ONLY_E2B_TOOLS frozenset is declared but never
used; update the E2B tool registration code that currently applies
_READONLY_ANNOTATION to all E2B tools unconditionally so it instead checks
membership in _READ_ONLY_E2B_TOOLS and only adds _READONLY_ANNOTATION for tools
whose names are in that set (or, if the intent is to mark all E2B tools
read-only, remove the unused _READ_ONLY_E2B_TOOLS constant). Specifically,
modify the E2B registration block that applies _READONLY_ANNOTATION to consult
_READ_ONLY_E2B_TOOLS when deciding to annotate each tool (referencing
_READ_ONLY_E2B_TOOLS and _READONLY_ANNOTATION to locate the change).
- Around line 450-453: The code currently applies _READONLY_ANNOTATION to every
entry from TOOL_REGISTRY; update the logic that builds the tool annotations (the
loop over TOOL_REGISTRY that adds _READONLY_ANNOTATION at/near where base_tool
is referenced) to only add that annotation when the tool's class property
BaseTool.read_only (accessed via base_tool.read_only) is truthy; in other words,
check base_tool.read_only before applying _READONLY_ANNOTATION so only tools
whose subclass overrides read_only=True (e.g., feature_requests, find_agent,
web_fetch, etc.) get readOnlyHint=True.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: d4584f49-d0d8-458c-8952-1ab04ecaf796
📒 Files selected for processing (1)
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.py
📜 Review details
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (13)
- GitHub Check: test (3.11)
- GitHub Check: end-to-end tests
- GitHub Check: get-changed-parts
- GitHub Check: setup
- GitHub Check: type-check (3.11)
- GitHub Check: test (3.13)
- GitHub Check: type-check (3.12)
- GitHub Check: type-check (3.13)
- GitHub Check: lint
- GitHub Check: Analyze (python)
- GitHub Check: Analyze (typescript)
- GitHub Check: check-overlaps
- GitHub Check: Seer Code Review
🧰 Additional context used
📓 Path-based instructions (2)
autogpt_platform/backend/**/*.py
📄 CodeRabbit inference engine (.github/copilot-instructions.md)
autogpt_platform/backend/**/*.py: Use Python 3.11 (required; managed by Poetry via pyproject.toml) for backend development
Always run 'poetry run format' (Black + isort) before linting in backend development
Always run 'poetry run lint' (ruff) after formatting in backend developmentRefer to
@backend/CLAUDE.mdfor backend-specific commands, architecture, and development tasks
autogpt_platform/backend/**/*.py: Import only at the top level; no local/inner imports except for lazy imports of heavy optional dependencies likeopenpyxl
Use absolute imports withfrom backend.module import ...for cross-package imports; single-dot relative imports (from .sibling import ...) are acceptable for sibling modules within the same package; avoid double-dot relative imports (from ..parent import ...)
Do not use duck typing withhasattr(),getattr(), orisinstance()for type dispatch; use typed interfaces, unions, or protocols instead
Use Pydantic models for structured data instead of dataclasses, namedtuples, or dicts
Do not use linter suppressors; no# type: ignore,# noqa, or# pyright: ignorecomments — fix the underlying type/code issue instead
Use list comprehensions instead of manual loop-and-append patterns
Use early return guard clauses to avoid deep nesting
Use%sfor deferred interpolation indebuglog statements; use f-strings for readability in other log levels (e.g.,logger.debug("Processing %s items", count),logger.info(f"Processing {count} items"))
Sanitize error paths usingos.path.basename()in error messages to avoid leaking directory structure
Avoid TOCTOU (time-of-check-time-of-use) patterns; do not use check-then-act patterns for file access and credit charging operations
Use Redis pipelines withtransaction=Truefor atomicity on multi-step Redis operations
Usemax(0, value)guards for computed values that should never be negative
Keep files under ~300 lines; if a file grows beyond this, split by responsibility (extract h...
Files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.py
autogpt_platform/**/*.py
📄 CodeRabbit inference engine (AGENTS.md)
Format Python code with
poetry run format
Files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.py
🧠 Learnings (8)
📓 Common learnings
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12356
File: autogpt_platform/backend/backend/copilot/constants.py:9-12
Timestamp: 2026-03-10T08:39:22.025Z
Learning: In Significant-Gravitas/AutoGPT PR `#12356`, the `COPILOT_SYNTHETIC_ID_PREFIX = "copilot-"` check in `create_auto_approval_record` (human_review.py) is intentional and safe. The `graph_exec_id` passed to this function comes from server-side `PendingHumanReview` DB records (not from user input); the API only accepts `node_exec_id` from users. Synthetic `copilot-*` IDs are only ever created server-side in `run_block.py`. The prefix skip avoids a DB lookup for a `AgentGraphExecution` record that legitimately does not exist for CoPilot sessions, while `user_id` scoping is enforced at the auth layer and on the resulting auto-approval record.
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12440
File: autogpt_platform/backend/backend/copilot/workflow_import/converter.py:0-0
Timestamp: 2026-03-17T10:57:12.953Z
Learning: In Significant-Gravitas/AutoGPT PR `#12440`, `autogpt_platform/backend/backend/copilot/workflow_import/converter.py` was fully rewritten (commit 732960e2d) to no longer make direct LLM/OpenAI API calls. The converter now builds a structured text prompt for AutoPilot/CoPilot instead. There is no `response.choices` access or any direct LLM client usage in this file. Do not flag `response.choices` access or LLM client initialization patterns as issues in this file.
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12439
File: autogpt_platform/backend/backend/blocks/autogpt_copilot.py:0-0
Timestamp: 2026-03-16T17:00:02.827Z
Learning: In autogpt_platform/backend/backend/blocks/autogpt_copilot.py, the recursion guard uses two module-level ContextVars: `_copilot_recursion_depth` (tracks current nesting depth) and `_copilot_recursion_limit` (stores the chain-wide ceiling). On the first invocation, `_copilot_recursion_limit` is set to `max_recursion_depth`; nested calls use `min(inherited_limit, max_recursion_depth)`, so they can only lower the cap, never raise it. The entry/exit logic is extracted into module-level helper functions. This is the approved pattern for preventing runaway sub-agent recursion in AutogptCopilotBlock (PR `#12439`, commits 348e9f8e2 and 3b70f61b1).
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12356
File: autogpt_platform/backend/backend/copilot/tools/run_block.py:297-300
Timestamp: 2026-03-10T08:38:33.249Z
Learning: In autogpt_platform/backend/backend/copilot/tools/run_block.py, the auto-approval key for sensitive block HITL review uses graph_exec_id (copilot-session-{session_id}) + node_id (copilot-node-{block_id}). This is intentional: approving a block type within a CoPilot session auto-approves all future invocations of that same block type within the same session, mirroring how auto-approve works in normal graph execution. The user explicitly opts into this session-scoped behavior via an auto-approve toggle. Without the toggle (default), each individual invocation requires its own approval.
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12213
File: autogpt_platform/frontend/src/app/api/openapi.json:9983-9995
Timestamp: 2026-02-27T15:59:00.370Z
Learning: Repo: Significant-Gravitas/AutoGPT PR: 12213 — OpenAPI/codegen
Learning: Ensuring a field is required in generated TS types needs two sides: (1) no default value on the Pydantic field, and (2) the OpenAPI model's "required" array must list it. For MCPToolInfo, making input_schema required in OpenAPI and removing Field(default_factory=dict) in the backend prevents optional typing drift.
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12356
File: autogpt_platform/backend/backend/copilot/tools/run_block.py:349-370
Timestamp: 2026-03-10T08:38:36.655Z
Learning: In the AutoGPT CoPilot HITL (Human-In-The-Loop) flow (`autogpt_platform/backend/backend/copilot/tools/run_block.py`), the review card presented to users sets `editable: false`, meaning reviewers cannot modify the input payload. Therefore, credentials resolved before `is_block_exec_need_review()` remain valid and do not need to be recomputed after the review step — the original `input_data` is unchanged through the review lifecycle.
Learnt from: Pwuts
Repo: Significant-Gravitas/AutoGPT PR: 12284
File: autogpt_platform/frontend/src/app/api/openapi.json:11897-11900
Timestamp: 2026-03-04T23:58:18.476Z
Learning: Repo: Significant-Gravitas/AutoGPT — PR `#12284`
Backend/frontend OpenAPI codegen convention: In backend/api/features/store/model.py, the StoreSubmission and StoreSubmissionAdminView models define submitted_at: datetime | None, changes_summary: str | None, and instructions: str | None with no default. This is intentional to produce “required but nullable” fields in OpenAPI (properties appear in required[] and use anyOf [type, null]). This matches Prisma’s submittedAt DateTime? and changesSummary String?. Do not flag this as a required/nullable mismatch.
Learnt from: Abhi1992002
Repo: Significant-Gravitas/AutoGPT PR: 12417
File: autogpt_platform/backend/backend/blocks/agent_mail/threads.py:80-102
Timestamp: 2026-03-16T16:30:20.657Z
Learning: In autogpt_platform/backend/backend/blocks/agent_mail/ blocks (and across the codebase), wrapping synchronous AgentMail SDK calls with `await asyncio.to_thread()` is NOT required. The block executor runs node execution in dedicated threads via `asyncio.run_coroutine_threadsafe` (manager.py lines ~745-752, ~1079), and the existing codebase pattern does not use `asyncio.to_thread` for SDK calls inside async `run()` methods.
📚 Learning: 2026-03-17T10:57:12.953Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12440
File: autogpt_platform/backend/backend/copilot/workflow_import/converter.py:0-0
Timestamp: 2026-03-17T10:57:12.953Z
Learning: In Significant-Gravitas/AutoGPT PR `#12440`, `autogpt_platform/backend/backend/copilot/workflow_import/converter.py` was fully rewritten (commit 732960e2d) to no longer make direct LLM/OpenAI API calls. The converter now builds a structured text prompt for AutoPilot/CoPilot instead. There is no `response.choices` access or any direct LLM client usage in this file. Do not flag `response.choices` access or LLM client initialization patterns as issues in this file.
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.py
📚 Learning: 2026-03-04T12:19:43.066Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12279
File: autogpt_platform/backend/backend/copilot/tools/base.py:184-188
Timestamp: 2026-03-04T12:19:43.066Z
Learning: In the AutoGPT Copilot backend (autogpt_platform/backend/backend/copilot/tools/), anonymous users always have user_id=None when passed to tool execution methods. The "anon_" prefix (e.g., "anon_123") is only used for PostHog/analytics tracking distinct_id and is never used as an actual user_id passed to tools. A simple truthiness check (`and user_id`) is sufficient to distinguish anonymous from authenticated users.
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.py
📚 Learning: 2026-02-26T17:02:22.448Z
Learnt from: Pwuts
Repo: Significant-Gravitas/AutoGPT PR: 12211
File: .pre-commit-config.yaml:160-179
Timestamp: 2026-02-26T17:02:22.448Z
Learning: Keep the pre-commit hook pattern broad for autogpt_platform/backend to ensure OpenAPI schema changes are captured. Do not narrow to backend/api/ alone, since the generated schema depends on Pydantic models across multiple directories (backend/data/, backend/blocks/, backend/copilot/, backend/integrations/, backend/util/). Narrowing could miss schema changes and cause frontend type desynchronization.
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.py
📚 Learning: 2026-03-04T08:04:35.881Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12273
File: autogpt_platform/backend/backend/copilot/tools/workspace_files.py:216-220
Timestamp: 2026-03-04T08:04:35.881Z
Learning: In the AutoGPT Copilot backend, ensure that SVG images are not treated as vision image types by excluding 'image/svg+xml' from INLINEABLE_MIME_TYPES and MULTIMODAL_TYPES in tool_adapter.py; the Claude API supports PNG, JPEG, GIF, and WebP for vision. SVGs (XML text) should be handled via the text path instead, not the vision path.
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.py
📚 Learning: 2026-03-05T15:42:08.207Z
Learnt from: ntindle
Repo: Significant-Gravitas/AutoGPT PR: 12297
File: .claude/skills/backend-check/SKILL.md:14-16
Timestamp: 2026-03-05T15:42:08.207Z
Learning: In Python files under autogpt_platform/backend (recursively), rely on poetry run format to perform formatting (Black + isort) and linting (ruff). Do not run poetry run lint as a separate step after poetry run format, since format already includes linting checks.
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.py
📚 Learning: 2026-03-16T16:35:40.236Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12440
File: autogpt_platform/backend/backend/api/features/workflow_import.py:54-63
Timestamp: 2026-03-16T16:35:40.236Z
Learning: Avoid using the word 'competitor' in public-facing identifiers and text. Use neutral naming for API paths, model names, function names, and UI text. Examples: rename 'CompetitorFormat' to 'SourcePlatform', 'convert_competitor_workflow' to 'convert_workflow', '/competitor-workflow' to '/workflow'. Apply this guideline to files under autogpt_platform/backend and autogpt_platform/frontend.
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.py
📚 Learning: 2026-03-31T15:37:38.626Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12623
File: autogpt_platform/backend/backend/copilot/tools/agent_generator/fixer.py:37-47
Timestamp: 2026-03-31T15:37:38.626Z
Learning: When validating/constructing Anthropic API model IDs in Significant-Gravitas/AutoGPT, allow the hyphen-separated Claude Opus 4.6 model ID `claude-opus-4-6` (it corresponds to `LlmModel.CLAUDE_4_6_OPUS` in `autogpt_platform/backend/backend/blocks/llm.py`). Do NOT require the dot-separated form in Anthropic contexts. Only OpenRouter routing variants should use the dot separator (e.g., `anthropic/claude-opus-4.6`); `claude-opus-4-6` should be treated as correct when passed to Anthropic, and flagged only if it’s used in the OpenRouter path where the dot form is expected.
Applied to files:
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.py
🔇 Additional comments (3)
autogpt_platform/backend/backend/copilot/sdk/tool_adapter.py (3)
17-17: LGTM — new import forToolAnnotations.The import from
mcp.typesis appropriate for the newreadOnlyHintannotation functionality.
444-445: LGTM — readonly annotation constant.Extracting
ToolAnnotations(readOnlyHint=True)as a module-level constant (_READONLY_ANNOTATION) avoids repeated allocations and clearly documents intent.
545-551: LGTM — Read tool marked as read-only.The
Readtool is correctly annotated withreadOnlyHint=Truesince reading files has no side effects.
Contributor
Author
|
/review |
…d _READ_ONLY_E2B_TOOLS Address review: since all tools now get readOnlyHint=True unconditionally, the BaseTool.read_only property and 15 tool class overrides were dead code. Remove them along with _READ_ONLY_E2B_TOOLS. Fix docstring to match.
Contributor
Author
|
🤖 🔵 Nit (Round 2): Comment in service.py line 1296 says "parallel execution of read-only tools" but all tools now have readOnlyHint. Should say "parallel execution of all tools". |
Contributor
Author
|
🤖 🔵 Nit (Round 3): |
…NNOTATION Address review nits: rename misleading constant (all tools get it, not just read-only), fix service.py comment to say "all tools".
github-actions
Bot
added
the
conflicts
Contributor
|
This pull request has conflicts with the base branch, please resolve those so we can evaluate the pull request. |
1 similar comment
Contributor
|
This pull request has conflicts with the base branch, please resolve those so we can evaluate the pull request. |
Keep readOnlyHint test, add SDK_DISALLOWED_TOOLS tests from dev.
github-actions
Bot
removed
the
conflicts
Contributor
|
Conflicts have been resolved! 🎉 A maintainer will review the pull request shortly. |
Contributor
Author
|
/review |
Reproduce the three production bugs that the pre-launch speculative execution mechanism caused, verifying the current direct-execution handler is free of them: Bug 1 (SECRT-2204): Duplicate execution on arg mismatch - test_single_execution_even_with_different_arg_representations - test_concurrent_calls_each_execute_once Bug 2: FIFO desync when security hook denies a tool - test_skipped_call_does_not_affect_subsequent_calls - test_handler_has_no_shared_queue_state Bug 3: Cancel race condition (task completes before cancel) - test_no_speculative_execution_before_handler_called - test_failed_execution_does_not_leave_orphaned_tasks
Each bug has two tests: one that reproduces the old buggy pre-launch behavior (xfail — proves the bug exists) and one that verifies the current clean handler is free of it (pass). Bug 1 (SECRT-2204) — duplicate execution on arg mismatch Bug 2 — FIFO desync when tool denied by security hook Bug 3 — cancel race (task completes before cancel arrives)
|
Contributor
Author
|
/review |
…e annotation test Add docstring note explaining why side-effect tools use readOnlyHint=True (deliberate override to avoid pre-launch duplicate-execution bug). Replace trivial ToolAnnotations construction test with assertion against the actual _PARALLEL_ANNOTATION constant.
![autogpt-pr-reviewer[bot]](https://avatars.githubusercontent.com/u/130738209?s=60&v=4){kind=small}
majdyz
commented
Apr 1, 2026
Contributor
Author
majdyz
left a comment
majdyz
left a comment
There was a problem hiding this comment.
Test Report (Round 2) — Local Verification
Branch: fix/copilot-duplicate-block-execution | Worktree: AutoGPT3
Unit Tests
poetry run pytest backend/copilot/sdk/tool_adapter_test.py -x -q
30 passed, 3 xfailed in 20.56s
All 30 tests pass. The 3 xfail tests are regression tests that reproduce the three old bugs inline and confirm they fail as expected.
Code Review Checklist
| Check | Result |
|---|---|
| Pre-launch infrastructure fully removed from production code | ✅ |
readOnlyHint=True annotation applied to all tools (TOOL_REGISTRY, E2B, Read) |
✅ |
No dangling references to removed functions (pre_launch_tool_call, cancel_pending_tool_tasks, _tool_task_queues) |
✅ |
All 4 cancel_pending_tool_tasks() calls removed from service.py |
✅ |
service.py imports clean (no removed symbols) |
✅ |
| Python import verification passes | ✅ |
tool_handler() executes exactly once per call (no speculative execution) |
✅ |
| Regression tests prove all 3 bugs (arg mismatch, FIFO desync, cancel race) | ✅ |
Verdict
PASS — Clean removal of pre-launch mechanism (-578 lines), replaced with SDK-native parallel dispatch via readOnlyHint annotations (+105 lines). Comprehensive regression tests included.
Abhi1992002
approved these changes
Apr 1, 2026
Contributor
Abhi1992002
left a comment
Abhi1992002
left a comment
There was a problem hiding this comment.
LGTM
- Codewise wise looks good
- Tested locally as well - tools are running in parallel perfectly
github-project-automation
Bot
moved this from 🆕 Needs initial review
to 👍🏼 Mergeable
in AutoGPT development kanban
github-project-automation
Bot
moved this from 👍🏼 Mergeable
to ✅ Done
in AutoGPT development kanban
4 tasks
8 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Why
CoPilot sessions are duplicating Linear tickets and GitHub PRs. Investigation of 5 production sessions (March 31st) found that 3/5 created duplicate Linear issues — each with consecutive IDs at the exact same timestamp, but only one visible in Langfuse traces.
Production gcloud logs confirm: 279 arg mismatch warnings per day, 37 duplicate block execution pairs, and all LinearCreateIssueBlock failures in pairs.
Related: SECRT-2204
What
Replace the speculative pre-launch mechanism with the SDK's native parallel dispatch via
readOnlyHinttool annotations. Remove ~580 lines of pre-launch infrastructure code.How
Root cause
The pre-launch mechanism had three compounding bugs:
AssistantMessage(used for pre-launch) and the MCPtools/calldispatch, causing frequent mismatches (279/day in prod)task.cancel()is best-effort in asyncio — if the HTTP call to Linear/GitHub already completed, the side effect is irreversibleFix
pre_launch_tool_call(),cancel_pending_tool_tasks(),_tool_task_queuesContextVar, all FIFO queue logic, and all 4cancel_pending_tool_tasks()calls inservice.pyreadOnlyHint=Trueannotations on 15+ read-only tools (find_block,search_docs,list_workspace_files, etc.) — the SDK CLI natively dispatches these in parallel (ref: anthropics/claude-code#14353)run_block,bash_exec,create_agent, etc.) have no annotation → CLI runs them sequentially → no duplicate execution riskNet change: -578 lines, +105 lines