Diagnostics and System Logs

The Diagnostics & Logs view provides real-time operational observability into background ingestion workers, Tree-sitter AST parsing tasks, webhook deliveries, and error traces.

---

Real-Time Log Ring Buffer

ContextCortex maintains an in-memory ring buffer of the latest 500 system events:

• Zero Disk Bloat: Diagnostic events are held in memory with strict bounding to prevent disk exhaustion.
• Log Level Filtering: Filter records by severity:
  • ALL: Complete operational stream.
  • INFO: Normal indexing and lifecycle updates.
  • WARNING: Recoverable warnings (e.g. rate limit thresholds, skipped binary files).
  • ERROR: Ingestion errors, failed clones, or unreachable databases.
  • DEBUG: Verbose chunking and Tree-sitter parser diagnostics.
• Keyword Filtering: Search message content and file paths interactively.
• Error Traceback Drawer: Click any ERROR log row to open an expanded slide-over drawer showing the full Python stack trace and offending parameters.
• Clear Buffer: Click Clear Logs to reset the in-memory buffer during testing sessions.

---

Health Check Endpoints

ContextCortex exposes standard health check endpoints for container orchestrators (Docker, Kubernetes, Kuma):

# Basic liveness probe
curl http://localhost:8021/health
# Response: {"status":"healthy"}

# Detailed diagnostic status
curl http://localhost:8021/admin/api/stats

---

Next Steps

• Explore dashboard theme palettes in Appearance and Theme Customization.
• Review architectural specifications in the System Architecture section.