Prompt file imported from intelligenceforgood/copilot (.github/prompts/pre-merge-review.prompt.md). Copyright stays with the author.

Pre-Merge Review

Execute the complete pre-merge review for all repos with changes.

Steps

1. Identify changed repos. Run git status -sb in each workspace root to find which repos have uncommitted or staged changes.

2. Read the full checklist. Load and follow the shared pre-merge checklist at copilot/.github/shared/pre-merge-checklist.instructions.md.

3. Code audit. Review all changed files against coding standards:

   • Type hints on every function (parameters + return)
   • Google-style docstrings on public/private methods
   • No unused imports, dead code, bare excepts, hard-coded secrets
   • Correct use of stores/factories/settings
   • No circular imports

4. Run quality gates for each changed repo:

   Python repos (core/, ssi/):

   cd <repo-root>
   conda run -n <env> pre-commit run --all-files   # Pass 1 — auto-fixes
   git add -u                                      # Stage fixes
   conda run -n <env> pre-commit run --all-files   # Pass 2 — must be clean
   

   Conda envs: i4g for core, i4g-ssi for ssi.

   UI repo:

   cd ui/ && make check && make build
   

   Infra repo:

   cd infra/ && terraform fmt -check -recursive
   

5. Run tests. conda run -n i4g pytest tests/unit in core (and ssi if changed).

6. Check docs/config. If behavior or env vars changed, verify:

   • docs/config/settings_manifest.yaml is updated
   • planning/change_log.md reflects the changes

7. Summary. Report: issues found, fixes applied, test results, hook output, remaining items.

8. Record lessons. If the review revealed new patterns, add to /memories/repo/lessons-learned.md.