shz-backend/.codex/skills/shz-pc-debug/references/project-topology.md

SHZ Project Topology

Workspace	Responsibility	Use during debugging
/Users/lapuda/code/shz-backend	Backend project	Investigate API routing, authorization, business services, scheduled work, persistence, and server errors. Use the project's log and guarded database skills here.
/Users/lapuda/code/shz-admin	PC frontend	Investigate PC job-seeker, supervisory-admin, and enterprise UI routes, components, client state, request construction, and browser interaction.
/Users/lapuda/code/shz-employment-service	Mini-program	Investigate only mini-program-specific behavior, or a suspected interface/authentication/business-logic concern shared with the PC application.

Select the minimum scope

1. Start a PC visual, interaction, navigation, or missing-request investigation in shz-admin after capturing browser evidence.
2. Start an API status, response, permission, server exception, or database-data investigation in shz-backend.
3. Investigate both shz-admin and shz-backend when the browser request or response shows that the failure crosses the client/server boundary.
4. Add shz-employment-service only after evidence indicates the same endpoint or business behavior is involved. Do not change mini-program code for a PC-only issue.

Work safely across repositories

1. Change into the selected workspace before using CodeGraph, running checks, or editing files.
2. Check each repository's working tree independently and preserve unrelated changes.
3. Attempt CodeGraph in the selected workspace before broad text search. If its index is unavailable, do not create one; use rg as the fallback.
4. Run focused verification in every repository changed. Do not deploy, restart services, or alter a remote environment unless the current user request explicitly authorizes committing and deploying; then follow deployment.md.