AI-Hub-Ghana/contributing

Canonical source for contribution docs used across `ai-hub-ghana` project repos. # contributing Canonical source for contribution docs used across `ai-hub-ghana` project repos. Edit here, sync out — don't hand-edit the copies inside individual project repos, they'll drift out of sync silently. ## What's here, and what isn't | File | Copied into | How it stays in sync | |---|---|---| | `docs/CLAUDE.md` | Every project repo (root) | Manual: `scripts/sync.sh` | | `docs/DESIGN_CONTRIBUTING.md` | Every project repo (root) | Manual: `scripts/sync.sh` | | — | `CONTRIBUTING.md` in `.github` | **Not duplicated here.** GitHub's own org-wide default mechanism already solves this natively — a repo without its own `CONTRIBUTING.md` automatically falls back to `.github`'s copy. Edit it directly at `ai-hub-ghana/.github`. | ## Updating a doc ```bash # 1. Edit the canonical file vim docs/CLAUDE.md # 2. Sync into each repo that needs it ./scripts/sync.sh ../website ./scripts/sync.sh ../repo-template # 3. In each target repo: review the diff, re-apply any # PROJECT-SPECIFIC sections (see note in CLAUDE.md), commit, PR. ``` The sync is a full overwrite, not a merge. `CLAUDE.md` has two sections marked `PROJECT-SPECIFIC` for exactly this reason — expect to reapply those by hand after a sync if a given repo has customized them. ## Why this isn't automated (yet) At 2-3 repos, a sync bot or scheduled Action to auto-open PRs is more infrastructure than the actual problem justifies — the manual script exists to prevent someone forgetting to update one of two or three copies, not to eliminate a five-minute task. Revisit automation once there are enough project repos that manual syncing becomes real toil, not before.