Documentation Updater

Project Overrides

!s="docs-updater"; for d in .specweave/skill-memories .claude/skill-memories "$HOME/.claude/skill-memories"; do p="$d/$s.md"; [ -f "$p" ] && awk '/^## Learnings$/{ok=1;next}/^## /{ok=0}ok' "$p" && break; done 2>/dev/null; true

Updates product documentation (.specweave/docs/) based on implementation progress.

When to Use

• Task specifies documentation updates in tasks.md
• Feature implementation is complete
• User says "update documentation" or "sync docs"
• After closing increment to ensure docs reflect reality

What It Does

1. Reads task requirements - Understands what was implemented from tasks.md
2. Updates living docs - Modifies .specweave/docs/ files with actual implementation
3. Status tracking - Changes [DRAFT] � [COMPLETE] on doc sections
4. Bidirectional links - Maintains links between docs and increments
5. Format adaptation - Matches existing doc structure (features/ or modules/)

Workflow

1. Read tasks.md � Find documentation tasks
2. Read implementation � Understand what changed
3. Update docs � Add real code examples, endpoints, configs
4. Mark complete � Change [DRAFT] to [COMPLETE]
5. Verify links � Ensure increment � doc references work

Example

tasks.md says:

**Documentation Updates**:
- [ ] .specweave/docs/features/payment.md [DRAFT]
- [ ] .specweave/docs/api/payments.md [DRAFT]

docs-updater does:

1. Reads payment implementation code
2. Updates payment.md with actual code examples
3. Updates API docs with real endpoints discovered in code
4. Changes status to [COMPLETE]

Integration Points

• Called by: spec-generator, task completion hooks
• Updates: .specweave/docs/**/*.md
• Reads: tasks.md, implementation code

Best Practices

• Run after completing feature tasks, not during
• Verify doc links are valid (use relative paths)
• Keep examples in sync with actual code
• Don't over-document - focus on what users need