Guides

Start with one coding-agent task. Open the next guide when you need it; a Python agent has its own secondary walkthrough.

• Python tool-calling agent
  • Add Maida to the project
  • Capture and review
  • Prove pass, fail, and repair
• Regression testing
  • Recommended workflow
  • Policy
  • Baseline contents
  • Reports and exit codes
  • Fail-fast and calibration
  • Reviewing changes
  • GitHub Actions
• Guardrails
  • What guardrails do
  • Available guardrails
  • LangChain / LangGraph
  • OpenAI Agents SDK
  • Quick examples
  • Configuration surfaces
  • How guardrails appear in traces
  • Choosing sensible defaults
  • Related docs
• Viewer
  • Usage
  • Development
• Capture Claude Code
  • Core capture-to-gate workflow
  • Import a captured session
  • Gate a capture locally
  • Run isolated scenarios
  • Command-hook fallback
• Scheduled checks
  • Run a scheduled check
  • Verdicts and scheduler behavior
  • Canary promotion
  • Input compatibility
• Gate draft extraction
  • Select workflows
  • Draft layout
  • Privacy and activation
• Local onboarding measurement
• Optional usage counts
  • Payload and consent
  • Prepared append-only receiver
  • What these counts establish