Repo Discovery Auditor
Inspect a repo before planning or coding.
Core workflow
- Start with high-signal files:
- package manager files
- app entrypoints and route layouts
- config files
- backend or data client setup
- representative feature pages or modules
- Identify:
- framework and runtime
- routing model
- state management
- data layer and auth
- UI system
- Map user-facing flows page by page or feature by feature.
- Look for maturity signals:
- validation
- loading and error states
- empty states
- access control
- reusable abstractions
- side-effect handling
- Look for risk signals:
- stale schema references
- duplicated business logic
- weak typing
- inconsistent auth or role checks
- debug code and ad-hoc branching
Evidence rule
Tie claims to actual files or patterns you inspected.
Do not claim runtime behavior you did not verify.
If something is inferred from static code only, say so.
Suggested output
Return results in this order when useful:
- Architecture summary
- Key stack and structure
- User-facing feature map
- What looks mature
- What looks rough or incomplete
- Important risks or inconsistencies
- Best next move
Maturity labels
Use simple bands:
- mature
- working but rough
- unclear or likely incomplete
Explain each label with concrete evidence.
Handoff use
If the repo audit will feed implementation work, end with a short section for Codex:
- likely files to touch
- architecture constraints
- risky areas to avoid breaking
- missing information that should be confirmed first