fix(platform): reduce Sentry noise by filtering expected errors and downgrading log levels#12430
Merged
fix(platform): reduce Sentry noise by filtering expected errors and downgrading log levels#12430
Conversation
…owngrading log levels - Add `before_send` hook to Sentry to filter known expected errors: AMQP/RabbitMQ connection errors, user credential errors (invalid API keys), insufficient balance, blocked IP access, Discord token errors, Google metadata DNS errors, inactive email recipients - Downgrade connection retry failures from error to warning in conn_retry - Stop sending user-caused ValueError exceptions to Sentry in block execution - Downgrade scheduler job failures and graph validation errors to warning - Wrap discord_system_alert to catch and warn instead of raising - Downgrade all notification system errors to warning (batch processing, email sending, user lookup, preference handling) - Downgrade cloud storage cleanup errors to warning - Downgrade executor shutdown/disconnect errors to warning
github-project-automation
Bot
moved this to 🆕 Needs initial review
in AutoGPT development kanban
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:
WalkthroughThis PR reduces Sentry noise and lowers log severity across executor, scheduler, notifications, retry, and utility modules; it adds a global Sentry before_send filter and prevents ValueError-derived validation/user exceptions from being sent to Sentry while preserving control flow and public APIs. Changes
Sequence Diagram(s)sequenceDiagram
participant Client as Caller
participant Scheduler as Scheduler
participant Executor as Executor/manager
participant Worker as Task/Block
participant Sentry as Sentry (before_send)
participant Notif as Notifications
Client->>Scheduler: submit / schedule job
Scheduler->>Executor: enqueue/run job
Executor->>Worker: execute node / run block
alt Block raises Validation (ValueError-derived)
Worker-->>Executor: raise ValidationError
Executor->>Sentry: skip capture (re-raise)
Executor->>Notif: log/warn
else Unexpected exception
Worker-->>Executor: raise unexpected exception
Executor->>Sentry: capture_exception(event)
Sentry-->>Sentry: _before_send filters/drops event
Sentry-->>Executor: delivered or suppressed
Executor->>Notif: notify (alert failures logged as warnings)
end
Scheduler->>Executor: unschedule/cleanup (warnings on validation/cleanup paths)
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~20 minutes Possibly related PRs
Suggested labels
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)
📝 Coding Plan
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 |
Contributor
🔍 PR Overlap DetectionThis check compares your PR against all other open PRs targeting the same branch to detect potential merge conflicts early. 🔴 Merge Conflicts DetectedThe following PRs have been tested and will have merge conflicts if merged after this PR. Consider coordinating with the authors.
🟢 Low Risk — File Overlap OnlyThese PRs touch the same files but different sections (click to expand)
Summary: 1 conflict(s), 0 medium risk, 2 low risk (out of 3 PRs with file overlap) Auto-generated on push. Ignores: |
![sentry[bot]](https://avatars.githubusercontent.com/in/12637?s=60&v=4){kind=small}
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
🧹 Nitpick comments (3)
autogpt_platform/backend/backend/util/retry.py (1)
184-186: Consider whether both log statements are needed.This callback logs the final failure, and then the exception propagates to
sync_wrapper/async_wrapper(lines 228/240) which log the same failure again. This results in duplicate warning logs for each final failure.If the duplication is intentional (different context levels), this is fine. Otherwise, consider removing one of the log statements.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@autogpt_platform/backend/backend/util/retry.py` around lines 184 - 186, The retry callback currently logs final failures via logger.warning when retry_state.outcome.failed and retry_state.next_action is None, which duplicates warnings already emitted later in sync_wrapper/async_wrapper; update the code to avoid duplicate logging by either removing the logger.warning call in the retry callback or lowering its level (e.g., to logger.debug) and keep the single warning in sync_wrapper/async_wrapper (functions referenced: retry_state, sync_wrapper, async_wrapper, and the retry callback block) so final failures are logged only once at the desired context/level.autogpt_platform/backend/backend/notifications/notifications.py (1)
1015-1018: Consider preserving traceback for debugging.The change from
logger.exceptiontologger.warningloses the traceback. Since the exception is re-raised and eventually caught by the@continuous_retrydecorator (which logs with full traceback every 10th failure), this is likely acceptable. However, for transient issues that resolve quickly, the traceback context might never be captured.If debugging becomes difficult, consider using
logger.warning(..., exc_info=True)to preserve the traceback while keeping the warning level.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@autogpt_platform/backend/backend/notifications/notifications.py` around lines 1015 - 1018, Update the except block so the log keeps the traceback while remaining at warning level: in the handler that currently does logger.warning(f"Error processing message in {queue_name}: {e}") (the block that then re-raises for message.process() and `@continuous_retry`), call logger.warning with exc_info=True (preserving the exception traceback) or otherwise include the exception info in the log call; keep the re-raise behavior unchanged.autogpt_platform/backend/backend/util/metrics.py (1)
24-88: Well-structured Sentry filter with comprehensive coverage.The filter correctly handles various expected error patterns. A few observations:
InsufficientBalanceError message coverage gap: The filter checks for
"insufficient balance"(line 54), but perautogpt_platform/backend/backend/util/exceptions.py:71-87,InsufficientBalanceErrorcan also have the message"You have no credits left to run an agent."which won't match this pattern.While
manager.pyseparately suppresses allValueErrorsubclasses (which includesInsufficientBalanceError), adding the alternate pattern here would provide defense-in-depth.Inactive email pattern (lines 72-73) requires both
"marked as inactive"AND"inactive addresses"in the message—this is quite specific and may miss some Postmark error variations.💡 Optional: Add alternate insufficient balance message pattern
# Expected business logic — insufficient balance - if "insufficient balance" in exc_msg: + if "insufficient balance" in exc_msg or "no credits left" in exc_msg: return None🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@autogpt_platform/backend/backend/util/metrics.py` around lines 24 - 88, Update the _before_send Sentry filter to catch the alternate InsufficientBalanceError message and broaden the bounced-email check: add the string "you have no credits left to run an agent." to the insufficient-balance matching (in the block that currently checks "insufficient balance") and change the inactive-email condition (currently requiring both "marked as inactive" AND "inactive addresses") to match if either phrase appears (use any/pattern list), so _before_send more robustly suppresses these known noisy errors.
🤖 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/notifications/notifications.py`:
- Around line 1015-1018: Update the except block so the log keeps the traceback
while remaining at warning level: in the handler that currently does
logger.warning(f"Error processing message in {queue_name}: {e}") (the block that
then re-raises for message.process() and `@continuous_retry`), call logger.warning
with exc_info=True (preserving the exception traceback) or otherwise include the
exception info in the log call; keep the re-raise behavior unchanged.
In `@autogpt_platform/backend/backend/util/metrics.py`:
- Around line 24-88: Update the _before_send Sentry filter to catch the
alternate InsufficientBalanceError message and broaden the bounced-email check:
add the string "you have no credits left to run an agent." to the
insufficient-balance matching (in the block that currently checks "insufficient
balance") and change the inactive-email condition (currently requiring both
"marked as inactive" AND "inactive addresses") to match if either phrase appears
(use any/pattern list), so _before_send more robustly suppresses these known
noisy errors.
In `@autogpt_platform/backend/backend/util/retry.py`:
- Around line 184-186: The retry callback currently logs final failures via
logger.warning when retry_state.outcome.failed and retry_state.next_action is
None, which duplicates warnings already emitted later in
sync_wrapper/async_wrapper; update the code to avoid duplicate logging by either
removing the logger.warning call in the retry callback or lowering its level
(e.g., to logger.debug) and keep the single warning in
sync_wrapper/async_wrapper (functions referenced: retry_state, sync_wrapper,
async_wrapper, and the retry callback block) so final failures are logged only
once at the desired context/level.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: 0920384f-ef26-4793-9277-54e3e58df9b0
📒 Files selected for processing (6)
autogpt_platform/backend/backend/executor/manager.pyautogpt_platform/backend/backend/executor/scheduler.pyautogpt_platform/backend/backend/notifications/notifications.pyautogpt_platform/backend/backend/util/cloud_storage.pyautogpt_platform/backend/backend/util/metrics.pyautogpt_platform/backend/backend/util/retry.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). (6)
- GitHub Check: types
- GitHub Check: Seer Code Review
- GitHub Check: test (3.11)
- GitHub Check: test (3.12)
- GitHub Check: test (3.13)
- 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 development
Files:
autogpt_platform/backend/backend/util/retry.pyautogpt_platform/backend/backend/util/metrics.pyautogpt_platform/backend/backend/notifications/notifications.pyautogpt_platform/backend/backend/executor/manager.pyautogpt_platform/backend/backend/executor/scheduler.pyautogpt_platform/backend/backend/util/cloud_storage.py
autogpt_platform/backend/**/*.{py,txt}
📄 CodeRabbit inference engine (autogpt_platform/backend/CLAUDE.md)
Use
poetry runprefix for all Python commands, including testing, linting, formatting, and migrations
Files:
autogpt_platform/backend/backend/util/retry.pyautogpt_platform/backend/backend/util/metrics.pyautogpt_platform/backend/backend/notifications/notifications.pyautogpt_platform/backend/backend/executor/manager.pyautogpt_platform/backend/backend/executor/scheduler.pyautogpt_platform/backend/backend/util/cloud_storage.py
autogpt_platform/backend/backend/**/*.py
📄 CodeRabbit inference engine (autogpt_platform/backend/CLAUDE.md)
Use Prisma ORM for database operations in PostgreSQL with pgvector for embeddings
Files:
autogpt_platform/backend/backend/util/retry.pyautogpt_platform/backend/backend/util/metrics.pyautogpt_platform/backend/backend/notifications/notifications.pyautogpt_platform/backend/backend/executor/manager.pyautogpt_platform/backend/backend/executor/scheduler.pyautogpt_platform/backend/backend/util/cloud_storage.py
autogpt_platform/**/*.py
📄 CodeRabbit inference engine (AGENTS.md)
Format Python code with
poetry run format
Files:
autogpt_platform/backend/backend/util/retry.pyautogpt_platform/backend/backend/util/metrics.pyautogpt_platform/backend/backend/notifications/notifications.pyautogpt_platform/backend/backend/executor/manager.pyautogpt_platform/backend/backend/executor/scheduler.pyautogpt_platform/backend/backend/util/cloud_storage.py
🧠 Learnings (4)
📚 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/util/retry.pyautogpt_platform/backend/backend/util/metrics.pyautogpt_platform/backend/backend/notifications/notifications.pyautogpt_platform/backend/backend/executor/manager.pyautogpt_platform/backend/backend/executor/scheduler.pyautogpt_platform/backend/backend/util/cloud_storage.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/util/retry.pyautogpt_platform/backend/backend/util/metrics.pyautogpt_platform/backend/backend/notifications/notifications.pyautogpt_platform/backend/backend/executor/manager.pyautogpt_platform/backend/backend/executor/scheduler.pyautogpt_platform/backend/backend/util/cloud_storage.py
📚 Learning: 2026-03-15T15:30:02.282Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12385
File: autogpt_platform/backend/backend/copilot/tools/helpers.py:149-185
Timestamp: 2026-03-15T15:30:02.282Z
Learning: In autogpt_platform/backend/backend/copilot/tools/helpers.py, within execute_block, when InsufficientBalanceError occurs after post-execution credit charging (concurrent balance drain after pre-check passed), this is treated as a non-fatal billing leak: log at ERROR level with structured JSON fields `{"billing_leak": True, "user_id": ..., "cost": ...}` for monitoring/alerting, then return BlockOutputResponse normally. Discarding the output would worsen UX since the block already executed with potential side effects. Reuse the credit_model obtained during the pre-execution balance check (guarded by `if cost > 0 and credit_model:`) for the post-execution charge; do not perform a second get_user_credit_model call.
Applied to files:
autogpt_platform/backend/backend/executor/manager.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/backend/blocks/**/*.py : Implement 'run' method with proper error handling in backend blocks
Applied to files:
autogpt_platform/backend/backend/executor/manager.py
🔇 Additional comments (18)
autogpt_platform/backend/backend/util/cloud_storage.py (1)
615-617: LGTM!Downgrading to
warningis appropriate here since cleanup failures are transient and recoverable—the operation will retry on the next cleanup cycle. The function already returns0gracefully, so this isn't a critical error path.autogpt_platform/backend/backend/executor/scheduler.py (3)
97-99: LGTM!Job failures logged by the listener are expected errors that will be handled by the job's error flow. Downgrading to
warningis appropriate since the scheduler continues operating normally.
139-141: LGTM!The exception is re-raised after logging, ensuring proper error propagation. Warning level is appropriate since the actual error handling occurs upstream.
188-202: LGTM!Graph validation failures are user-caused errors (e.g., invalid graph configuration). Downgrading to
warningis appropriate since these are expected operational scenarios, not platform bugs.autogpt_platform/backend/backend/util/retry.py (2)
66-68: LGTM!Discord alert failures are secondary to the primary error being handled. Warning level is appropriate since the alert is a best-effort notification.
227-241: LGTM on the log level change.Downgrading to
warningis consistent with the PR's noise reduction goal. The exception is still re-raised, so error handling continues normally.autogpt_platform/backend/backend/util/metrics.py (1)
105-105: LGTM!The
before_sendhook is correctly wired into Sentry initialization.autogpt_platform/backend/backend/executor/manager.py (5)
377-385: LGTM! Good separation of user errors from platform errors.Suppressing
ValueErrorsubclasses (BlockExecutionError,BlockInputError,InsufficientBalanceError) from Sentry is appropriate since these represent expected user/validation errors, not platform bugs. The exception is still re-raised to maintain normal error flow.
1483-1485: LGTM!Discord alert failures are secondary to the primary operation. Warning level is appropriate.
1908-1916: LGTM!Graceful shutdown scenarios where threads don't finish in time or disconnect fails are operational concerns, not critical errors. Warning level is appropriate since the service is intentionally stopping.
1931-1935: LGTM!Errors during shutdown signaling are expected edge cases. Warning level is appropriate.
1964-1976: LGTM!Both the timeout warning and executor shutdown errors are graceful degradation scenarios during service shutdown. Warning level is appropriate.
autogpt_platform/backend/backend/notifications/notifications.py (6)
418-421: Good defensive change.Wrapping
discord_send_alertin a try/except prevents Discord API failures from crashing the notification service or interrupting other operations. This is critical for service resilience.
306-310: LGTM on batch processing log level changes.These scenarios (concurrent batch clearing, missing user emails, empty batch data, parsing errors) are expected operational states, not platform bugs. Warning level is appropriate since processing continues gracefully.
Also applies to: 321-324, 347-356, 375-378
521-554: LGTM!Summary data gathering failures returning sensible defaults with a warning is a good fallback pattern. Users still receive a summary (with empty data) rather than no notification at all.
568-572: LGTM!Consistent with other batch processing paths—concurrent clearing is an expected race condition scenario.
620-622: LGTM!Missing user emails are expected scenarios (deleted users, unverified emails). Warning level with graceful return/continue is appropriate.
Also applies to: 657-659, 678-680, 954-956
752-754: LGTM on all batch chunk processing error paths.These handle various email API failure scenarios (size limits, inactive recipients, malformed data). Warning level is appropriate since the code has proper remediation logic (removing bad notifications, disabling users, etc.).
Also applies to: 777-782, 796-798, 830-857, 866-870, 884-886, 892-903, 914-918
- Make "connection refused" filter AMQP-specific by checking exception module/type instead of broadly matching all ConnectionRefusedError - Add "no credits left" as alternate insufficient balance pattern - Broaden inactive email filter (OR instead of AND) - Remove duplicate final-failure log in conn_retry on_retry callback - Preserve traceback with exc_info=True in notification queue consumer
Contributor
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 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/util/metrics.py`:
- Around line 88-96: The current noisy_patterns check unconditionally drops any
log with "connection refused", which can hide non-AMQP outages; change the logic
around noisy_patterns/msg so that "connection refused" is only filtered when the
message is AMQP-related (e.g., contains "amqp", "amqpconnection",
"connection_forced" or other AMQP-specific tokens), mirroring the AMQP-only
scoping used in the exception branch; update the check that builds
noisy_patterns or add an additional conditional so only AMQP-related "connection
refused" messages return None while other "connection refused" logs are
preserved.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: 62c69319-7740-46f1-86a6-835208193c1b
📒 Files selected for processing (3)
autogpt_platform/backend/backend/notifications/notifications.pyautogpt_platform/backend/backend/util/metrics.pyautogpt_platform/backend/backend/util/retry.py
🚧 Files skipped from review as they are similar to previous changes (2)
- autogpt_platform/backend/backend/util/retry.py
- autogpt_platform/backend/backend/notifications/notifications.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). (6)
- GitHub Check: types
- GitHub Check: Seer Code Review
- GitHub Check: test (3.13)
- GitHub Check: test (3.12)
- GitHub Check: test (3.11)
- 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 development
Files:
autogpt_platform/backend/backend/util/metrics.py
autogpt_platform/backend/**/*.{py,txt}
📄 CodeRabbit inference engine (autogpt_platform/backend/CLAUDE.md)
Use
poetry runprefix for all Python commands, including testing, linting, formatting, and migrations
Files:
autogpt_platform/backend/backend/util/metrics.py
autogpt_platform/backend/backend/**/*.py
📄 CodeRabbit inference engine (autogpt_platform/backend/CLAUDE.md)
Use Prisma ORM for database operations in PostgreSQL with pgvector for embeddings
Files:
autogpt_platform/backend/backend/util/metrics.py
autogpt_platform/**/*.py
📄 CodeRabbit inference engine (AGENTS.md)
Format Python code with
poetry run format
Files:
autogpt_platform/backend/backend/util/metrics.py
🧠 Learnings (3)
📓 Common learnings
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12356
File: autogpt_platform/backend/backend/copilot/constants.py:9-12
Timestamp: 2026-03-10T08:39:13.707Z
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.
📚 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/util/metrics.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/util/metrics.py
🔇 Additional comments (1)
autogpt_platform/backend/backend/util/metrics.py (1)
41-50: Good AMQP-specific narrowing for exception-sideconnection refused.This is a solid safeguard: it reduces noisy RabbitMQ errors without suppressing unrelated service failures.
- Use explicit exception types instead of broad isinstance(ValueError) in execute_node to avoid suppressing platform errors like NotFoundError - Make metadata.google.internal filter GCP-aware (only filter in non-cloud envs) - Scope log-side "connection refused" filter to AMQP-related messages only
Tests were checking mock_logger.error but the log calls were changed to logger.warning. Update assertions to check mock_logger.warning instead.
Sentry's LoggingIntegration stores log messages under event["logentry"], not event["message"]. The previous check was dead code that never matched.
Contributor
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 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/executor/manager.py`:
- Around line 389-400: The current exception whitelist omits plain ValueError
causing expected user/input ValueErrors raised by execute_node to be treated as
unexpected; update the isinstance check around variable ex (the block starting
with if not isinstance(ex, (...))) to include ValueError in the tuple so
ValueError instances from execute_node (referenced at execute_node lines that
raise ValueError) are treated as expected and not sent to Sentry.
In `@autogpt_platform/backend/backend/util/metrics.py`:
- Around line 89-90: The log-noise filter only reads event["message"], missing
Sentry's rendered text at event["logentry"]["formatted"]; update the logic where
event and msg are handled (the block using event.get("logger") and
event.get("message") and the variable msg) to also check event.get("logentry",
{}).get("formatted") and prefer that value (lowercased) when present, falling
back to event["message"] so Sentry LoggingIntegration formatted messages are
filtered the same way as regular messages.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: 42740d0d-3c87-401c-a6dc-af162188240e
📒 Files selected for processing (3)
autogpt_platform/backend/backend/executor/manager.pyautogpt_platform/backend/backend/notifications/test_notifications.pyautogpt_platform/backend/backend/util/metrics.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). (5)
- GitHub Check: types
- GitHub Check: test (3.12)
- GitHub Check: test (3.13)
- GitHub Check: test (3.11)
- GitHub Check: Check PR Status
🧰 Additional context used
📓 Path-based instructions (5)
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 development
Files:
autogpt_platform/backend/backend/notifications/test_notifications.pyautogpt_platform/backend/backend/util/metrics.pyautogpt_platform/backend/backend/executor/manager.py
autogpt_platform/backend/**/*.{py,txt}
📄 CodeRabbit inference engine (autogpt_platform/backend/CLAUDE.md)
Use
poetry runprefix for all Python commands, including testing, linting, formatting, and migrations
Files:
autogpt_platform/backend/backend/notifications/test_notifications.pyautogpt_platform/backend/backend/util/metrics.pyautogpt_platform/backend/backend/executor/manager.py
autogpt_platform/backend/backend/**/*.py
📄 CodeRabbit inference engine (autogpt_platform/backend/CLAUDE.md)
Use Prisma ORM for database operations in PostgreSQL with pgvector for embeddings
Files:
autogpt_platform/backend/backend/notifications/test_notifications.pyautogpt_platform/backend/backend/util/metrics.pyautogpt_platform/backend/backend/executor/manager.py
autogpt_platform/**/*.py
📄 CodeRabbit inference engine (AGENTS.md)
Format Python code with
poetry run format
Files:
autogpt_platform/backend/backend/notifications/test_notifications.pyautogpt_platform/backend/backend/util/metrics.pyautogpt_platform/backend/backend/executor/manager.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/notifications/test_notifications.py
🧠 Learnings (16)
📓 Common learnings
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12356
File: autogpt_platform/backend/backend/copilot/constants.py:9-12
Timestamp: 2026-03-10T08:39:13.707Z
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: 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.
📚 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/notifications/test_notifications.py
📚 Learning: 2026-02-04T16:50:20.508Z
Learnt from: CR
Repo: Significant-Gravitas/AutoGPT PR: 0
File: autogpt_platform/backend/CLAUDE.md:0-0
Timestamp: 2026-02-04T16:50:20.508Z
Learning: Applies to autogpt_platform/backend/**/*_test.py : Always review snapshot changes with `git diff` before committing when updating snapshots with `poetry run pytest --snapshot-update`
Applied to files:
autogpt_platform/backend/backend/notifications/test_notifications.pyautogpt_platform/backend/backend/util/metrics.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/notifications/test_notifications.pyautogpt_platform/backend/backend/util/metrics.pyautogpt_platform/backend/backend/executor/manager.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/notifications/test_notifications.pyautogpt_platform/backend/backend/util/metrics.pyautogpt_platform/backend/backend/executor/manager.py
📚 Learning: 2026-03-10T08:39:13.707Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12356
File: autogpt_platform/backend/backend/copilot/constants.py:9-12
Timestamp: 2026-03-10T08:39:13.707Z
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.
Applied to files:
autogpt_platform/backend/backend/util/metrics.py
📚 Learning: 2026-03-04T23:58:18.476Z
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.
Applied to files:
autogpt_platform/backend/backend/util/metrics.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/util/metrics.py
📚 Learning: 2026-03-10T08:38:30.834Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12356
File: autogpt_platform/backend/backend/copilot/tools/run_block.py:349-370
Timestamp: 2026-03-10T08:38:30.834Z
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.
Applied to files:
autogpt_platform/backend/backend/util/metrics.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/util/metrics.py
📚 Learning: 2026-03-11T16:50:37.459Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12367
File: autogpt_platform/backend/generate_views.py:146-150
Timestamp: 2026-03-11T16:50:37.459Z
Learning: In Significant-Gravitas/AutoGPT PR `#12367` (`autogpt_platform/backend/generate_views.py`), the `security_invoker = false` setting on analytics views is intentional. `analytics_readonly` is granted access only to the `analytics` schema. The views execute as their owner (postgres) to read `auth.*` and `platform.*` tables, but expose only curated, scrubbed columns via their SELECT lists. This is the correct PostgreSQL "view as security boundary" pattern — equivalent to a stored-procedure security definer. Do not flag this as a privilege escalation; switching to `security_invoker = true` would require granting `analytics_readonly` direct SELECT on `auth.sessions`, `auth.audit_log_entries`, `platform.AgentGraphExecution`, etc., which is a much wider blast radius than the current design.
Applied to files:
autogpt_platform/backend/backend/util/metrics.py
📚 Learning: 2026-03-15T23:15:58.145Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12398
File: autogpt_platform/backend/scripts/measure_copilot_tokens.py:0-0
Timestamp: 2026-03-15T23:15:58.145Z
Learning: In Significant-Gravitas/AutoGPT CoPilot, developer token-measurement scripts (e.g. `scripts/measure_copilot_tokens.py`) intentionally use a simplified, hardcoded base prompt rather than the production prompt-construction path. This isolates tool-schema token cost from the variable production system prompt, making before/after token diffs solely attributable to schema changes. Do not flag this as a drift risk in such utility scripts.
Applied to files:
autogpt_platform/backend/backend/util/metrics.py
📚 Learning: 2026-03-15T16:52:11.575Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12426
File: autogpt_platform/backend/backend/copilot/sdk/service.py:0-0
Timestamp: 2026-03-15T16:52:11.575Z
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.
Applied to files:
autogpt_platform/backend/backend/util/metrics.py
📚 Learning: 2026-03-01T07:59:02.311Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12213
File: autogpt_platform/frontend/src/app/api/openapi.json:10030-10037
Timestamp: 2026-03-01T07:59:02.311Z
Learning: Repo: Significant-Gravitas/AutoGPT PR: 12213 — For MCP manual token storage, backend model autogpt_platform/backend/backend/api/features/mcp/routes.py defines MCPStoreTokenRequest.token as Pydantic SecretStr with a min length constraint, which generates OpenAPI schema metadata (format: "password", writeOnly: true, minLength: 1) in autogpt_platform/frontend/src/app/api/openapi.json. Prefer SecretStr (with length constraints) for sensitive request fields so generated TS clients and docs treat them as secrets.
Applied to files:
autogpt_platform/backend/backend/util/metrics.py
📚 Learning: 2026-03-15T15:30:02.282Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12385
File: autogpt_platform/backend/backend/copilot/tools/helpers.py:149-185
Timestamp: 2026-03-15T15:30:02.282Z
Learning: In autogpt_platform/backend/backend/copilot/tools/helpers.py, within execute_block, when InsufficientBalanceError occurs after post-execution credit charging (concurrent balance drain after pre-check passed), this is treated as a non-fatal billing leak: log at ERROR level with structured JSON fields `{"billing_leak": True, "user_id": ..., "cost": ...}` for monitoring/alerting, then return BlockOutputResponse normally. Discarding the output would worsen UX since the block already executed with potential side effects. Reuse the credit_model obtained during the pre-execution balance check (guarded by `if cost > 0 and credit_model:`) for the post-execution charge; do not perform a second get_user_credit_model call.
Applied to files:
autogpt_platform/backend/backend/executor/manager.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/backend/blocks/**/*.py : Implement 'run' method with proper error handling in backend blocks
Applied to files:
autogpt_platform/backend/backend/executor/manager.py
🔇 Additional comments (3)
autogpt_platform/backend/backend/executor/manager.py (1)
1503-1503: Log-level downgrades are consistent with non-fatal shutdown/alert flows.This warning-level treatment looks appropriate for operationally expected cleanup/notification failures.
Also applies to: 1928-1935, 1951-1953, 1984-1986, 1995-1995
autogpt_platform/backend/backend/util/metrics.py (1)
77-82: Good call on environment-scoped metadata filtering and hook wiring.The non-cloud guard for
metadata.google.internalandbefore_send=_before_sendintegration are aligned and reduce false-positive suppression risk.Also applies to: 122-122
autogpt_platform/backend/backend/notifications/test_notifications.py (1)
259-262: Test expectation updates are correct for warning-level logging.These assertions now match the downgraded warning behavior and keep coverage on the same error-handling branches.
Also applies to: 374-378, 481-485
execute_node raises plain ValueError for disabled blocks and expired/missing auth. Use isinstance(ex, ValueError) but exclude NotFoundError and GraphNotFoundError which could indicate real platform issues.
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 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/executor/manager.py`:
- Around line 1918-1920: The current timeout handling around
thread.join(timeout=300) is incorrect because join() doesn't raise TimeoutError;
replace the try/except block with a join(timeout=300) followed by an explicit
check of thread.is_alive() to detect a timeout, and if still alive call the
existing disconnect/cleanup logic and log the warning (referencing the same
logger.warning call and prefix used), removing the unused TimeoutError except
branch; ensure any forced disconnect uses the same disconnect function/method
invoked elsewhere in this scope.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: 686e040e-3045-4317-8476-2b9d1feca382
📒 Files selected for processing (1)
autogpt_platform/backend/backend/executor/manager.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). (6)
- GitHub Check: types
- GitHub Check: test (3.11)
- GitHub Check: test (3.12)
- GitHub Check: test (3.13)
- GitHub Check: Seer Code Review
- 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 development
Files:
autogpt_platform/backend/backend/executor/manager.py
autogpt_platform/backend/**/*.{py,txt}
📄 CodeRabbit inference engine (autogpt_platform/backend/CLAUDE.md)
Use
poetry runprefix for all Python commands, including testing, linting, formatting, and migrations
Files:
autogpt_platform/backend/backend/executor/manager.py
autogpt_platform/backend/backend/**/*.py
📄 CodeRabbit inference engine (autogpt_platform/backend/CLAUDE.md)
Use Prisma ORM for database operations in PostgreSQL with pgvector for embeddings
Files:
autogpt_platform/backend/backend/executor/manager.py
autogpt_platform/**/*.py
📄 CodeRabbit inference engine (AGENTS.md)
Format Python code with
poetry run format
Files:
autogpt_platform/backend/backend/executor/manager.py
🧠 Learnings (7)
📚 Learning: 2026-03-15T15:30:02.282Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12385
File: autogpt_platform/backend/backend/copilot/tools/helpers.py:149-185
Timestamp: 2026-03-15T15:30:02.282Z
Learning: In autogpt_platform/backend/backend/copilot/tools/helpers.py, within execute_block, when InsufficientBalanceError occurs after post-execution credit charging (concurrent balance drain after pre-check passed), this is treated as a non-fatal billing leak: log at ERROR level with structured JSON fields `{"billing_leak": True, "user_id": ..., "cost": ...}` for monitoring/alerting, then return BlockOutputResponse normally. Discarding the output would worsen UX since the block already executed with potential side effects. Reuse the credit_model obtained during the pre-execution balance check (guarded by `if cost > 0 and credit_model:`) for the post-execution charge; do not perform a second get_user_credit_model call.
Applied to files:
autogpt_platform/backend/backend/executor/manager.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/backend/blocks/**/*.py : Implement 'run' method with proper error handling in backend blocks
Applied to files:
autogpt_platform/backend/backend/executor/manager.py
📚 Learning: 2026-03-10T08:39:13.707Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12356
File: autogpt_platform/backend/backend/copilot/constants.py:9-12
Timestamp: 2026-03-10T08:39:13.707Z
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.
Applied to files:
autogpt_platform/backend/backend/executor/manager.py
📚 Learning: 2026-03-10T08:38:30.834Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12356
File: autogpt_platform/backend/backend/copilot/tools/run_block.py:349-370
Timestamp: 2026-03-10T08:38:30.834Z
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.
Applied to files:
autogpt_platform/backend/backend/executor/manager.py
📚 Learning: 2026-03-10T08:38:29.078Z
Learnt from: majdyz
Repo: Significant-Gravitas/AutoGPT PR: 12356
File: autogpt_platform/backend/backend/copilot/tools/run_block.py:297-300
Timestamp: 2026-03-10T08:38:29.078Z
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.
Applied to files:
autogpt_platform/backend/backend/executor/manager.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/executor/manager.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/executor/manager.py
🔇 Additional comments (2)
autogpt_platform/backend/backend/executor/manager.py (2)
64-69: Good exception-noise boundary for Sentry capture.The Line 387-Line 391 logic is a solid balance: expected
ValueErrorpaths are suppressed whileNotFoundError/GraphNotFoundErrorstill get captured.Also applies to: 383-392
1493-1493: Log-level downgrades are appropriate for cleanup/notification paths.These are non-user-actionable/transient paths, so warning level is a better fit than error.
Also applies to: 1925-1925, 1941-1943, 1974-1976, 1985-1985
thread.join(timeout=N) never raises TimeoutError in Python — it always returns None. Use thread.is_alive() after join to detect timeout.
8 tasks
Swiftyos
approved these changes
Mar 16, 2026
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
Contributor
Author
|
Addressed all 3 nitpick comments from the initial review:
|
This was referenced Mar 23, 2026
Bentlybro
pushed a commit
that referenced
this pull request
Apr 4, 2026
…owngrading log levels (#12430) ## Summary Reduces Sentry error noise by ~90% by filtering out expected/transient errors and downgrading inappropriate error-level logs to warnings. Most of the top Sentry issues are not actual bugs but expected conditions (user errors, transient infra, business logic) that were incorrectly logged at ERROR level, causing them to be captured as Sentry events. ## Changes ### 1. Sentry `before_send` filter (`metrics.py`) Added a `before_send` hook to filter known expected errors before they reach Sentry: - **AMQP/RabbitMQ connection errors** — transient during deploys/restarts - **User credential errors** — invalid API keys, missing auth headers (user error, not platform bug) - **Insufficient balance** — expected business logic - **Blocked IP access** — security check working as intended - **Discord bot token errors** — misconfiguration, not runtime error - **Google metadata DNS errors** — expected in non-GCP environments - **Inactive email recipients** — expected for bounced addresses - **Unclosed client sessions/connectors** — resource cleanup noise ### 2. Connection retry log levels (`retry.py`) - `conn_retry` final failure: `error` → `warning` (these are infra retries, not bugs) - `conn_retry` wrapper final failure: `error` → `warning` - Discord alert send failure: `error` → `warning` ### 3. Block execution Sentry capture (`manager.py`) - Skip `sentry_sdk.capture_exception()` for `ValueError` subclasses (BlockExecutionError, BlockInputError, InsufficientBalanceError, etc.) — these are user-caused errors, not platform bugs - Downgrade executor shutdown/disconnect errors to warning ### 4. Scheduler log levels (`scheduler.py`) - Graph validation failure: `error` → `warning` (expected for old/invalid graphs) - Unable to unschedule graph: `error` → `warning` - Job listener failure: `error` → `warning` - Async operation failure: `error` → `warning` ### 5. Discord system alert (`notifications.py`) - Wrapped `discord_system_alert` endpoint with try/catch to prevent unhandled exceptions (fixes AUTOGPT-SERVER-743, AUTOGPT-SERVER-7MW) ### 6. Notification system log levels (`notifications.py`) - All batch processing errors: `error` → `warning` - User email not found: `error` → `warning` - Notification parsing errors: `error` → `warning` - Email sending failures: `error` → `warning` - Summary data gathering failure: `error` → `warning` - Cleaned up unprofessional error messages ### 7. Cloud storage cleanup (`cloud_storage.py`) - Cleanup error: `error` → `warning` ## Sentry Issues Addressed ### AMQP/RabbitMQ (~3.4M events total) - [AUTOGPT-SERVER-3H2](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-3H2) — AMQPConnector ConnectionRefusedError (1.2M events) - [AUTOGPT-SERVER-3H3](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-3H3) — AMQPConnectionWorkflowFailed (770K events) - [AUTOGPT-SERVER-3H4](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-3H4) — AMQP connection workflow failed (770K events) - [AUTOGPT-SERVER-3H5](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-3H5) — AMQPConnectionWorkflow reporting failure (770K events) - [AUTOGPT-SERVER-3H7](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-3H7) — Socket failed to connect (514K events) - [AUTOGPT-SERVER-3H8](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-3H8) — TCP Connection attempt failed (514K events) - [AUTOGPT-SERVER-3H6](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-3H6) — AMQPConnectionError (93K events) - [AUTOGPT-SERVER-7SX](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-7SX) — Error creating transport (69K events) - [AUTOGPT-SERVER-1TN](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-1TN) — ChannelInvalidStateError (39K events) - [AUTOGPT-SERVER-6JC](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-6JC) — ConnectionClosedByBroker (2K events) - [AUTOGPT-SERVER-6RJ/6RK/6RN/6RQ/6RP/6RR](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-6RJ) — Various connection failures (~15K events) - [AUTOGPT-SERVER-4A5/6RM/7XN](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-4A5) — Connection close/transport errors (~540 events) ### User Credential Errors (~15K events) - [AUTOGPT-SERVER-6S5](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-6S5) — Incorrect OpenAI API key (9.2K events) - [AUTOGPT-SERVER-7W4](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-7W4) — Incorrect API key in AIConditionBlock (3.4K events) - [AUTOGPT-SERVER-83Y](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-83Y) — AI condition invalid key (2.3K events) - [AUTOGPT-SERVER-7ZP](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-7ZP) — Perplexity missing auth header (451 events) - [AUTOGPT-SERVER-7XK/7XM](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-7XK) — Anthropic invalid key (125 events) - [AUTOGPT-SERVER-82C](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-82C) — Missing auth header (27 events) - [AUTOGPT-SERVER-721](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-721) — Ideogram invalid token (165 events) ### Business Logic / Validation (~120K events) - [AUTOGPT-SERVER-7YQ](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-7YQ) — Disabled block used in graph (56K events) - [AUTOGPT-SERVER-6W3](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-6W3) — Graph failed validation (46K events) - [AUTOGPT-SERVER-6W2](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-6W2) — Unable to unschedule graph (46K events) - [AUTOGPT-SERVER-83X](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-83X) — Blocked IP access (15K events) - [AUTOGPT-SERVER-6K9](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-6K9) — Insufficient balance (4K events) ### Discord Alert Failures (~24K events) - [AUTOGPT-SERVER-743](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-743) — Discord improper token (22K events) - [AUTOGPT-SERVER-7MW](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-7MW) — Discord 403 Missing Access (1.5K events) ### Notification System (~16K events) - [AUTOGPT-SERVER-550](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-550) — Notification batch create error (8.3K events) - [AUTOGPT-SERVER-58H](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-58H) — ValidationError for NotificationEventModel (3K events) - [AUTOGPT-SERVER-5C6](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-5C6) — Get notification batch error (2.1K events) - [AUTOGPT-SERVER-4BT](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-4BT) — Notification batch create error (1.8K events) - [AUTOGPT-SERVER-5E4](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-5E4) — NotificationPreference validation (1.4K events) - [AUTOGPT-SERVER-508](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-508) — Inactive email recipients (702 events) ### Infrastructure / Transient (~20K events) - [AUTOGPT-SERVER-6WJ](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-6WJ) — Unclosed client session (13K events) - [AUTOGPT-SERVER-745](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-745) — Unclosed connector (5.8K events) - [AUTOGPT-SERVER-4V1](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-4V1) — Google metadata DNS error (2.2K events) - [AUTOGPT-SERVER-80J](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-80J) — CloudStorage DNS error (35 events) ### Executor Shutdown - [AUTOGPT-SERVER-55J](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-55J) — Error disconnecting run client (118 events) ## Test plan - [x] All pre-commit hooks pass (Ruff, isort, Black, Pyright typecheck) - [x] All changed modules import successfully - [ ] Deploy to staging and verify Sentry event volume drops significantly - [ ] Verify legitimate errors still appear in Sentry
Bentlybro
pushed a commit
to Bentlybro/AutoGPT
that referenced
this pull request
Apr 4, 2026
…owngrading log levels (Significant-Gravitas#12430) ## Summary Reduces Sentry error noise by ~90% by filtering out expected/transient errors and downgrading inappropriate error-level logs to warnings. Most of the top Sentry issues are not actual bugs but expected conditions (user errors, transient infra, business logic) that were incorrectly logged at ERROR level, causing them to be captured as Sentry events. ## Changes ### 1. Sentry `before_send` filter (`metrics.py`) Added a `before_send` hook to filter known expected errors before they reach Sentry: - **AMQP/RabbitMQ connection errors** — transient during deploys/restarts - **User credential errors** — invalid API keys, missing auth headers (user error, not platform bug) - **Insufficient balance** — expected business logic - **Blocked IP access** — security check working as intended - **Discord bot token errors** — misconfiguration, not runtime error - **Google metadata DNS errors** — expected in non-GCP environments - **Inactive email recipients** — expected for bounced addresses - **Unclosed client sessions/connectors** — resource cleanup noise ### 2. Connection retry log levels (`retry.py`) - `conn_retry` final failure: `error` → `warning` (these are infra retries, not bugs) - `conn_retry` wrapper final failure: `error` → `warning` - Discord alert send failure: `error` → `warning` ### 3. Block execution Sentry capture (`manager.py`) - Skip `sentry_sdk.capture_exception()` for `ValueError` subclasses (BlockExecutionError, BlockInputError, InsufficientBalanceError, etc.) — these are user-caused errors, not platform bugs - Downgrade executor shutdown/disconnect errors to warning ### 4. Scheduler log levels (`scheduler.py`) - Graph validation failure: `error` → `warning` (expected for old/invalid graphs) - Unable to unschedule graph: `error` → `warning` - Job listener failure: `error` → `warning` - Async operation failure: `error` → `warning` ### 5. Discord system alert (`notifications.py`) - Wrapped `discord_system_alert` endpoint with try/catch to prevent unhandled exceptions (fixes AUTOGPT-SERVER-743, AUTOGPT-SERVER-7MW) ### 6. Notification system log levels (`notifications.py`) - All batch processing errors: `error` → `warning` - User email not found: `error` → `warning` - Notification parsing errors: `error` → `warning` - Email sending failures: `error` → `warning` - Summary data gathering failure: `error` → `warning` - Cleaned up unprofessional error messages ### 7. Cloud storage cleanup (`cloud_storage.py`) - Cleanup error: `error` → `warning` ## Sentry Issues Addressed ### AMQP/RabbitMQ (~3.4M events total) - [AUTOGPT-SERVER-3H2](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-3H2) — AMQPConnector ConnectionRefusedError (1.2M events) - [AUTOGPT-SERVER-3H3](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-3H3) — AMQPConnectionWorkflowFailed (770K events) - [AUTOGPT-SERVER-3H4](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-3H4) — AMQP connection workflow failed (770K events) - [AUTOGPT-SERVER-3H5](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-3H5) — AMQPConnectionWorkflow reporting failure (770K events) - [AUTOGPT-SERVER-3H7](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-3H7) — Socket failed to connect (514K events) - [AUTOGPT-SERVER-3H8](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-3H8) — TCP Connection attempt failed (514K events) - [AUTOGPT-SERVER-3H6](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-3H6) — AMQPConnectionError (93K events) - [AUTOGPT-SERVER-7SX](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-7SX) — Error creating transport (69K events) - [AUTOGPT-SERVER-1TN](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-1TN) — ChannelInvalidStateError (39K events) - [AUTOGPT-SERVER-6JC](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-6JC) — ConnectionClosedByBroker (2K events) - [AUTOGPT-SERVER-6RJ/6RK/6RN/6RQ/6RP/6RR](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-6RJ) — Various connection failures (~15K events) - [AUTOGPT-SERVER-4A5/6RM/7XN](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-4A5) — Connection close/transport errors (~540 events) ### User Credential Errors (~15K events) - [AUTOGPT-SERVER-6S5](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-6S5) — Incorrect OpenAI API key (9.2K events) - [AUTOGPT-SERVER-7W4](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-7W4) — Incorrect API key in AIConditionBlock (3.4K events) - [AUTOGPT-SERVER-83Y](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-83Y) — AI condition invalid key (2.3K events) - [AUTOGPT-SERVER-7ZP](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-7ZP) — Perplexity missing auth header (451 events) - [AUTOGPT-SERVER-7XK/7XM](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-7XK) — Anthropic invalid key (125 events) - [AUTOGPT-SERVER-82C](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-82C) — Missing auth header (27 events) - [AUTOGPT-SERVER-721](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-721) — Ideogram invalid token (165 events) ### Business Logic / Validation (~120K events) - [AUTOGPT-SERVER-7YQ](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-7YQ) — Disabled block used in graph (56K events) - [AUTOGPT-SERVER-6W3](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-6W3) — Graph failed validation (46K events) - [AUTOGPT-SERVER-6W2](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-6W2) — Unable to unschedule graph (46K events) - [AUTOGPT-SERVER-83X](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-83X) — Blocked IP access (15K events) - [AUTOGPT-SERVER-6K9](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-6K9) — Insufficient balance (4K events) ### Discord Alert Failures (~24K events) - [AUTOGPT-SERVER-743](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-743) — Discord improper token (22K events) - [AUTOGPT-SERVER-7MW](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-7MW) — Discord 403 Missing Access (1.5K events) ### Notification System (~16K events) - [AUTOGPT-SERVER-550](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-550) — Notification batch create error (8.3K events) - [AUTOGPT-SERVER-58H](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-58H) — ValidationError for NotificationEventModel (3K events) - [AUTOGPT-SERVER-5C6](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-5C6) — Get notification batch error (2.1K events) - [AUTOGPT-SERVER-4BT](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-4BT) — Notification batch create error (1.8K events) - [AUTOGPT-SERVER-5E4](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-5E4) — NotificationPreference validation (1.4K events) - [AUTOGPT-SERVER-508](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-508) — Inactive email recipients (702 events) ### Infrastructure / Transient (~20K events) - [AUTOGPT-SERVER-6WJ](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-6WJ) — Unclosed client session (13K events) - [AUTOGPT-SERVER-745](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-745) — Unclosed connector (5.8K events) - [AUTOGPT-SERVER-4V1](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-4V1) — Google metadata DNS error (2.2K events) - [AUTOGPT-SERVER-80J](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-80J) — CloudStorage DNS error (35 events) ### Executor Shutdown - [AUTOGPT-SERVER-55J](https://significant-gravitas.sentry.io/issues/AUTOGPT-SERVER-55J) — Error disconnecting run client (118 events) ## Test plan - [x] All pre-commit hooks pass (Ruff, isort, Black, Pyright typecheck) - [x] All changed modules import successfully - [ ] Deploy to staging and verify Sentry event volume drops significantly - [ ] Verify legitimate errors still appear in Sentry
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.
Summary
Reduces Sentry error noise by ~90% by filtering out expected/transient errors and downgrading inappropriate error-level logs to warnings. Most of the top Sentry issues are not actual bugs but expected conditions (user errors, transient infra, business logic) that were incorrectly logged at ERROR level, causing them to be captured as Sentry events.
Changes
1. Sentry
before_sendfilter (metrics.py)Added a
before_sendhook to filter known expected errors before they reach Sentry:2. Connection retry log levels (
retry.py)conn_retryfinal failure:error→warning(these are infra retries, not bugs)conn_retrywrapper final failure:error→warningerror→warning3. Block execution Sentry capture (
manager.py)sentry_sdk.capture_exception()forValueErrorsubclasses (BlockExecutionError, BlockInputError, InsufficientBalanceError, etc.) — these are user-caused errors, not platform bugs4. Scheduler log levels (
scheduler.py)error→warning(expected for old/invalid graphs)error→warningerror→warningerror→warning5. Discord system alert (
notifications.py)discord_system_alertendpoint with try/catch to prevent unhandled exceptions (fixes AUTOGPT-SERVER-743, AUTOGPT-SERVER-7MW)6. Notification system log levels (
notifications.py)error→warningerror→warningerror→warningerror→warningerror→warning7. Cloud storage cleanup (
cloud_storage.py)error→warningSentry Issues Addressed
AMQP/RabbitMQ (~3.4M events total)
User Credential Errors (~15K events)
Business Logic / Validation (~120K events)
Discord Alert Failures (~24K events)
Notification System (~16K events)
Infrastructure / Transient (~20K events)
Executor Shutdown
Test plan