feat(logging): Add structured logging (Issue #695) by anchapin · Pull Request #736 · anchapin/ModPorter-AI

anchapin · 2026-03-07T20:04:42Z

Summary

Implements GitHub Issue #695: Add structured logging

Changes Made:

Add structlog>=24.0.0 to backend and ai-engine requirements
Implement structlog-based structured logging in backend
Add configure_structlog() function with JSON format support
Add LoggingMiddleware for request/response logging with correlation IDs
Add RequestContextMiddleware for context management
Update ai-engine logging with structlog support
Configure JSON format for production environments (auto-detected from ENVIRONMENT=production)
Add correlation ID support for distributed tracing

Features:

Structured JSON Logging: Logs are output in JSON format in production
Correlation ID Support: Each request gets a unique correlation ID for tracing
Request/Response Logging: Middleware logs all HTTP requests and responses
Context Management: Context variables are properly propagated across async calls

Environment Variables:

: Set to "true" to enable JSON format (auto-enabled in production)
: Enable debug mode for verbose console output
: When set to "production", auto-enables JSON logging

Co-authored-by: openhands openhands@all-hands.dev

Implements GitHub Issue #700 This CODEOWNERS file defines code ownership and review requirements for the ModPorter-AI project: - Frontend: /frontend/ directory ownership - Backend: /backend/ directory ownership - AI-Engine: /ai-engine/ directory ownership - Infrastructure: Docker and docker-compose files - Security: Security-related scripts and configs - Documentation: docs/ and markdown files - Configuration: Project-wide configs and CI/CD workflows Readiness Pillar: Security Co-authored-by: openhands <openhands@all-hands.dev>

Implement comprehensive health check endpoints for Kubernetes probes: - Add /health/readiness endpoint with dependency checks (DB, Redis) - Add /health/liveness endpoint to verify process is running - Add /health basic health check endpoint - Include latency metrics for each dependency - Support degraded status when non-critical dependencies fail Readiness Pillar: Debugging & Observability Co-authored-by: openhands <openhands@all-hands.dev>

- Add sentry-sdk to backend with FastAPI and SQLAlchemy integrations - Add @sentry/react to frontend with browser tracing and replay - Integrate Sentry with existing ErrorBoundary component - Add Sentry configuration to .env.example files - Configure environment variables for DSN and sampling rates - Add performance monitoring and error capturing This implements the Readiness Pillar for Debugging & Observability by providing comprehensive error tracking for production issue detection. Co-authored-by: openhands <openhands@all-hands.dev>

- Add structlog dependency to backend and ai-engine requirements - Implement structlog-based structured logging in backend - Add configure_structlog() function with JSON format support - Add LoggingMiddleware for request/response logging with correlation IDs - Add RequestContextMiddleware for context management - Update ai-engine logging with structlog support - Configure JSON format for production environments (auto-detected) - Add correlation ID support for distributed tracing Co-authored-by: openhands <openhands@all-hands.dev>

sourcery-ai

Sorry @anchapin, you have reached your weekly rate limit of 500000 diff characters.

Please try again later or upgrade to continue using Sourcery

Copilot

Pull request overview

This PR aims to improve observability across the stack by introducing structured logging (structlog), correlation IDs, and request/response logging middleware, and it also adds health endpoints and Sentry configuration updates.

Changes:

Add structlog-based structured logging configuration to backend and AI engine, including correlation ID context binding.
Introduce backend request/response logging middleware and add new Kubernetes-style health endpoints (/health, /health/liveness, /health/readiness).
Add Sentry support/configuration in the frontend (and Sentry SDK dependency in the backend).

Reviewed changes

Copilot reviewed 16 out of 16 changed files in this pull request and generated 17 comments.

Show a summary per file

File	Description
frontend/src/main.tsx	Initializes Sentry in the frontend and configures tracing/replay.
frontend/src/components/ErrorBoundary/ErrorBoundary.tsx	Reports caught errors to Sentry.
frontend/package.json	Adds `@sentry/react`.
frontend/.env.example	Documents Sentry-related frontend env vars.
backend/src/services/structured_logging.py	Adds structlog configuration and context binding helpers.
backend/src/services/logging_middleware.py	New middleware for request/response logging + request context setup.
backend/src/main.py	Wires in logging middleware, structlog configuration, and health router.
backend/src/api/health.py	Adds liveness/readiness/basic health endpoints with DB/Redis checks.
backend/requirements.txt	Adds `structlog` and `sentry-sdk[fastapi]`.
backend/main.py	Initializes Sentry in the legacy backend entrypoint.
ai-engine/utils/logging_config.py	Adds structlog configuration + correlation ID helpers for AI engine.
ai-engine/main.py	Calls structlog configuration during AI engine startup.
ai-engine/requirements.txt	Adds `structlog`.
README.md	Updates health-check documentation and adds endpoint details.
.github/CODEOWNERS	Adds code ownership rules.
.env.example	Documents backend Sentry sampling rate and version env vars.

Comments suppressed due to low confidence (1)

ai-engine/main.py:36

configure_structlog() configures/clears root logging handlers, but setup_logging() immediately clears and reconfigures them again. As written, this likely negates the intended JSON/structlog formatting (structlog-rendered JSON will be wrapped by AgentLogFormatter). Consider consolidating into a single logging configuration path so production logs are consistently structured.

# Also configure structlog for structured JSON logging in production
configure_structlog(
    debug_mode=debug_mode,
    json_format=os.getenv("LOG_JSON_FORMAT", "false").lower() == "true"
)

setup_logging(
    debug_mode=debug_mode,
    enable_file_logging=os.getenv("ENABLE_FILE_LOGGING", "true").lower() == "true"
)

Copilot · 2026-03-07T20:15:03Z

backend/src/main.py

+# Health check endpoints (no prefix - used for Kubernetes probes)
+app.include_router(health.router)


This PR is titled/described as adding structured logging (Issue #695), but this change also introduces new backend health endpoints (/health, /health/readiness, /health/liveness) and documents them. Please confirm this scope is intentional and update the PR description/title (or split into a separate PR/issue) so reviewers can track and approve the additional behavior change appropriately.

Copilot · 2026-03-07T20:15:03Z

frontend/src/main.tsx

+    replaysOnErrorSampleRate: 1.0,
+    // Filter events
+    beforeSend(event) {
+      // Don't send events in development unless explicitly enabled
+      if (


VITE_SENTRY_ENABLE_DEV is a string in Vite env, so !import.meta.env.VITE_SENTRY_ENABLE_DEV only checks whether it is defined, not whether it is set to "true"/"false". This makes VITE_SENTRY_ENABLE_DEV="false" still enable Sentry in development; parse the env var to a boolean explicitly (e.g., compare lowercased string to "true").

Copilot · 2026-03-07T20:15:03Z

backend/src/services/structured_logging.py

    correlation_id_var.set(correlation_id)
+    structlog.contextvars.clear_contextvars()
+    structlog.contextvars.bind_contextvars(correlation_id=correlation_id)
    return correlation_id


set_correlation_id() now clears and binds structlog contextvars, but clear_correlation_id() (later in this module) only resets the local ContextVar and does not clear structlog’s context. This can leak a previous correlation_id into subsequent logs/requests; update the clear path to also clear/unbind structlog contextvars.

Copilot · 2026-03-07T20:15:04Z

backend/src/services/structured_logging.py

        metadata: Dictionary of metadata to store
    """
    request_metadata_var.set(metadata)
+    structlog.contextvars.bind_contextvars(**metadata)



set_request_metadata() now binds metadata into structlog contextvars, but clear_request_metadata() only clears the local request_metadata_var. Without unbinding/clearing structlog contextvars, request fields can leak into later logs. Ensure the clear path also clears structlog contextvars (or restores prior context).

Copilot · 2026-03-07T20:15:04Z

backend/src/api/health.py

+- /health/readiness: Checks if the application can serve traffic (dependencies available)
+- /health/liveness: Checks if the application is running and doesn't need to be restarted
+
+Issue #699: Add health check endpoints


Module docstring references “Issue #699”, but this PR/feature is for Issue #695. Please align the issue reference so it’s clear which change introduced these endpoints (or remove the issue number if it’s not accurate).

Suggested change

Issue #699: Add health check endpoints

Issue #695: Add health check endpoints

Copilot · 2026-03-07T20:15:05Z

backend/src/services/logging_middleware.py

+        # Generate correlation ID for this request
+        correlation_id = set_correlation_id()


LoggingMiddleware always calls set_correlation_id() to generate a new ID, which overwrites any incoming X-Correlation-ID that RequestContextMiddleware may have set. To support distributed tracing, prefer get_correlation_id() and only generate a new ID if none is already present (and ensure the response header echoes the effective ID).

Suggested change

# Generate correlation ID for this request

correlation_id = set_correlation_id()

# Get existing correlation ID (if set by upstream middleware) or generate a new one

existing_correlation_id = get_correlation_id()

correlation_id = existing_correlation_id or set_correlation_id()

Copilot · 2026-03-07T20:15:06Z

backend/src/services/logging_middleware.py

+        # Get correlation ID from header if present
+        correlation_id = request.headers.get("X-Correlation-ID")
+        if correlation_id:
+            set_correlation_id(correlation_id)
+


RequestContextMiddleware sets a correlation ID from the incoming header but never clears it in the finally block. Because contextvars are per-task but can be reused across requests in the same worker, this risks correlation IDs leaking into later requests (especially for paths excluded from LoggingMiddleware). Call clear_correlation_id() in the finally block.

Copilot · 2026-03-07T20:15:06Z

backend/src/services/logging_middleware.py

+class LoggingMiddleware(BaseHTTPMiddleware):
+    """
+    Middleware for logging HTTP requests and responses with correlation IDs.
+
+    Features:
+    - Automatic correlation ID generation for each request
+    - Request/response timing
+    - Structured logging of HTTP method, path, status code
+    - Request/response body size logging
+    """


New request/response logging and correlation-ID behavior is introduced here, but there are no corresponding tests verifying (a) X-Correlation-ID propagation from incoming headers, (b) response headers are set, and (c) context is cleared between requests. The backend has a substantial test suite; please add unit/integration tests covering these middleware behaviors.

Copilot · 2026-03-07T20:15:06Z

frontend/src/main.tsx

+        return null;
+      }
+      return event;
+    },


Sentry.addCaptureConsoleIntegration() is not part of the @sentry/react public API in recent SDKs and will throw at runtime / fail type-checking. If the intent is to capture console output, add the capture-console integration via Sentry.init({ integrations: [...] }) (or remove this line).

Copilot · 2026-03-07T20:15:06Z

README.md

+| Endpoint | Purpose | Dependencies Checked |
+|----------|---------|---------------------|
+| `/health` | Basic health check | None |
+| `/health/liveness` | Process is running | None |
+| `/health/readiness` | Can serve traffic | Database, Redis |


The health-check endpoints table is malformed Markdown because each row starts with || instead of |. This won’t render as a table on GitHub; remove the extra leading pipe so it’s a standard markdown table.

Suggested change

| Endpoint | Purpose | Dependencies Checked |

|----------|---------|---------------------|

| `/health` | Basic health check | None |

| `/health/liveness` | Process is running | None |

| `/health/readiness` | Can serve traffic | Database, Redis |

| Endpoint | Purpose | Dependencies Checked |

|--------------------|---------------------|----------------------|

| `/health` | Basic health check | None |

| `/health/liveness` | Process is running | None |

| `/health/readiness`| Can serve traffic | Database, Redis |

openhands-agent added 4 commits March 7, 2026 14:54

Copilot AI review requested due to automatic review settings March 7, 2026 20:04

sourcery-ai bot reviewed Mar 7, 2026

View reviewed changes

Copilot started reviewing on behalf of anchapin March 7, 2026 20:06 View session

fix: Apply prettier formatting to src/main.tsx

d943f79

Copilot AI reviewed Mar 7, 2026

View reviewed changes

openhands-agent added 2 commits March 7, 2026 15:24

fix: Update pnpm lockfile with missing @sentry/react

a82f866

fix: Apply ruff formatting to backend files

d5b1769

anchapin merged commit 20f1f1b into main Mar 7, 2026
21 of 24 checks passed

anchapin deleted the feature/issue-695-structured-logging branch March 7, 2026 21:08

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat(logging): Add structured logging (Issue #695)#736

feat(logging): Add structured logging (Issue #695)#736
anchapin merged 7 commits intomainfrom
feature/issue-695-structured-logging

anchapin commented Mar 7, 2026

Uh oh!

sourcery-ai bot left a comment

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Mar 7, 2026

Uh oh!

Copilot AI Mar 7, 2026

Uh oh!

Copilot AI Mar 7, 2026

Uh oh!

Copilot AI Mar 7, 2026

Uh oh!

Copilot AI Mar 7, 2026

Uh oh!

Copilot AI Mar 7, 2026

Uh oh!

Copilot AI Mar 7, 2026

Uh oh!

Copilot AI Mar 7, 2026

Uh oh!

Copilot AI Mar 7, 2026

Uh oh!

Copilot AI Mar 7, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

		# Health check endpoints (no prefix - used for Kubernetes probes)
		app.include_router(health.router)

	Issue #699: Add health check endpoints
	Issue #695: Add health check endpoints

		# Generate correlation ID for this request
		correlation_id = set_correlation_id()

-        # Generate correlation ID for this request
-        correlation_id = set_correlation_id()
+        # Get existing correlation ID (if set by upstream middleware) or generate a new one
+        existing_correlation_id = get_correlation_id()
+        correlation_id = existing_correlation_id or set_correlation_id()

Conversation

anchapin commented Mar 7, 2026

Summary

Changes Made:

Features:

Environment Variables:

Uh oh!

sourcery-ai bot left a comment

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Mar 7, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 7, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 7, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 7, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 7, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 7, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 7, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 7, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 7, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 7, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants