7. Operations and reference

Running SPEAR day to day: health checks, known failures and their fixes, the test suites, and where every file lives.

• 7.1. Operations
  • 7.1.1. Health check
  • 7.1.2. Reading a failure
  • 7.1.3. Inspecting a live scope
  • 7.1.4. Cleaning up after an interrupted session
  • 7.1.5. Switching model
  • 7.1.6. Logs and audit
  • 7.1.7. Known environment quirks
• 7.2. Troubleshooting
  • 7.2.1. No corpus index
  • 7.2.2. Backend unavailable
  • 7.2.3. Requests refused whole
  • 7.2.4. No standard bound
  • 7.2.5. The answer was withheld
  • 7.2.6. Stale or missing evidence
  • 7.2.7. Build or test failure during agent work
  • 7.2.8. Network disabled
  • 7.2.9. Wrong writable root
  • 7.2.10. Where to look next
• 7.3. Testing
  • 7.3.1. Running the suites
  • 7.3.2. Opt-in suites
  • 7.3.3. What the suites assert
  • 7.3.4. Determinism
  • 7.3.5. Residual state
  • 7.3.6. Live smoke test
• 7.4. Directory layout
  • 7.4.1. Top level
  • 7.4.2. The application
  • 7.4.3. Configuration
  • 7.4.4. Content the deployment owns
  • 7.4.5. Persistent state
  • 7.4.6. Entry points
  • 7.4.7. This documentation
  • 7.4.8. Building