codd-assemble
Assemble generated CoDD sprint fragments into a complete, buildable project. Use after all `codd implement` sprints have produced…
Reverse-propagate source code changes back to affected CoDD design documents. Use after code changes when the user needs to map changed source files to modules, identify impacted design docs, and optionally update those docs while preserving CoDD coherence.
$ npx -y skills add yohey-w/codd-dev --skill codd-propagate --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/codd-propagateContext preview
The summary Claude sees to decide when to auto-load this skill.
Reverse-propagate source code changes back to affected CoDD design documents. Use after code changes when the user needs to map changed source files to modules, identify impacted design docs, and optionally update those docs while preserving CoDD coherence.
name: codd-propagate description: | Reverse-propagate source code changes back to affected CoDD design documents. Use after code changes when the user needs to map changed source files to modules, identify impacted design docs, and optionally update those docs while preserving CoDD coherence.
Reverse-propagate source code changes to design documents. When source code changes, `codd propagate` identifies which design docs are affected via the `modules` field and optionally updates them using AI.
Do NOT use this for:
CoDD supports two propagation directions:
| Direction | Command | Trigger | Mechanism | |-----------|---------|---------|-----------| | Forward (doc → doc) | `codd impact` | Design doc changed | CEG dependency graph (`depends_on`) | | Reverse (code → doc) | `codd propagate` | Source code changed | `modules` field in frontmatter |
`/codd-propagate` handles the **reverse** direction only.
git diff HEAD # 1. Detect changed source files ↓ source_dirs mapping # 2. Map files to modules (e.g. src/auth/service.py → "auth") ↓ modules field lookup # 3. Find design docs with matching modules ↓ Analysis report # 4. Show which docs need attention ↓ (--update only) AI update # 5. AI updates affected doc bodies
Before running propagate, verify:
1. `codd/codd.yaml` exists with `source_dirs` configured in `scan`. 2. Design docs have `modules` field in their frontmatter. 3. There are uncommitted source code changes (or specify `--diff` target). 4. Confirm you are in the intended project root.
If `modules` fields are missing, run `codd plan --init` to regenerate wave_config with modules, then regenerate or restore design docs.
codd propagate --path .
This shows:
Review the analysis before deciding whether to update.
codd propagate --path . --update
This additionally:
After update, review every changed document before committing.
1. **After analysis**: Review the file → module → doc mapping. Confirm the affected docs list is correct before running with `--update`. 2. **After update**: Read every updated design doc. Confirm:
Not every code change needs a design doc update. The AI prompt explicitly tells the model to leave the body unchanged for:
Use analysis-only mode (`codd propagate` without `--update`) to check which docs are affected, then decide manually whether an update is warranted.
| Flag | Description | |------|-------------| | `--path` | Project root (default: current directory) | | `--diff` | Git diff target (default: `HEAD`) | | `--update` | Actually update affected docs via AI (default: analysis only) | | `--ai-cmd` | Override AI command for this run |
After modifying source code:
# 1. Check reverse impact (code → docs) codd propagate --path . # 2. If design docs need updating codd propagate --path . --update # 3. Rebuild dependency graph codd scan --path . # 4. Check forward impact (doc → doc) if design docs changed codd impact --path . # 5. Validate and commit codd validate --path . git add -A && git commit
CoDD: Coherence-Driven Development — Claude Code plugin for cross-artifact change impact analysis
Assemble generated CoDD sprint fragments into a complete, buildable project. Use after all `codd implement` sprints have produced…
Conversationally evolve an existing CoDD project. Use when the user describes a functional change in natural language ("add…
Consult Claude Fable 5 (Anthropic's most capable model for deep, long-horizon reasoning — or whichever model currently holds that…
Generate CoDD design documents for a greenfield project one wave at a time. Use after `codd init` and prepared requirements when…
Run the CoDD greenfield autopilot: a requirements document in, a working system out, unattended. Use when the user wants to build…
Analyze the downstream impact of changed requirements, design docs, code, or tests in a CoDD project. Use when the user needs to…