xiduzo/wtf / skills/wtf.setup
wtf.setup
VerifiedThis skill should be used when a user wants to set up WTF in a repository or check that the environment is ready for it — for example "set up wtf", "is wtf ready", "install wtf templates", or "a new dev joined, set them up". Verifies the GitHub CLI and its auth, installs the required gh extensions and the issue and PR templates, configures issue classification, planning, feature scope, and delivery mode, and registers the hooks. Run once per repo when onboarding, or when a contributor joins the project.
shared-references/steering-doc-process.md
View on GitHub# Steering Document Process
Shared procedure for every `wtf.steer-*` skill (`steer-vision`, `steer-tech`, `steer-design`, `steer-qa`). Each skill inherits this structure. Only the doc name, research checklist, gap-topic list, and template vary.
Every steering doc lives at `docs/steering/<NAME>.md`. Treat the docs as living. Generate once from research and interview. Then refine. Never regenerate from scratch.
## Consumer-side load
Any non-steer skill that needs steering context follows this single procedure. Skills citing it: `wtf.design-feature`, `wtf.design-trace`, `wtf.implement-trace`, `wtf.hotfix`, `wtf.pr-review`, `wtf.spike`, `wtf.verify-trace`, `wtf.retro`, `wtf.reflect`.
1. Use the Read tool to attempt reading `docs/steering/<DOC>.md` (one of `TECH.md`, `QA.md`, `DESIGN.md`, `VISION.md`).
2. If it **exists**, keep its content in context. Apply it silently throughout the session. Do not surface it to the user.
3. If it **does not exist**, choose the appropriate behavior for the skill:
- **Strict consumer** (skill cannot do its job without the doc — for example `wtf.design-trace`, `wtf.design-feature`, `wtf.implement-trace`, `wtf.verify-trace`): apply `./questioning-style.md` and ask "`docs/steering/<DOC>.md` does not exist yet. <one-line description of what the doc captures>. Would you like to create it now?" — header `<Doc> steering doc missing`:
- **Create it now** → invoke the matching `wtf.steer-<doc>` skill (recommended), then return and continue
- **Skip for this session** → continue without it. Decisions will not reference project standards.
- **Best-effort consumer** (skill can degrade gracefully — for example `wtf.hotfix`, `wtf.pr-review`, `wtf.spike`): silently continue without it. Note in any output if the doc would have changed the recommendation.
| Skill | Doc(s) consumed | Mode |
|---|---|---|
| `wtf.implement-trace` | `TECH.md`, `QA.md` (coverage threshold) | strict for TECH, soft default for QA |
| `wtf.design-trace` | `DESIGN.md` | strict |
| `wtf.design-feature` | `DESIGN.md` | strict |
| `wtf.verify-trace` | `QA.md` | strict |
| `wtf.pr-review` | `TECH.md` | best-effort |
| `wtf.hotfix` | `TECH.md` | best-effort |
| `wtf.spike` | `TECH.md` | best-effort |
| `wtf.retro` | `TECH.md`, `QA.md`, `DESIGN.md`, `VISION.md` | best-effort, route deltas |
| `wtf.reflect` | all four (writes back) | producer — see "Hard-Won Lessons" below |
## Hard-Won Lessons (writer-side, for `wtf.reflect`)
`wtf.reflect` writes session learnings into the consuming docs under a `## Hard-Won Lessons` section. The bullet format and routing rules live in `wtf.reflect` itself. The steering docs accept the appended bullets without further structure changes.
## Step 1. Check if the document already exists
Use the Read tool to attempt reading `docs/steering/<DOC>.md`.
If the file **exists**, apply `./questioning-style.md` and ask "docs/steering/<DOC>.md already exists. What would you like to do?" — header `<Doc> found`:
- **Refine it** → read the current doc, then skip to step 4. Use the existing doc as context. Only ask about gaps or outdated sections.
- **Exit** → leave it as-is and exit immediately.
If the file does not exist, continue to step 2.
## Step 2. Research the codebase
Use the Agent tool to extract facts directly from the codebase. Do not ask the user for things that can be read. Each `steer-*` skill owns its own research checklist. See the skill's SKILL.md.
Synthesize findings internally. Produce a concrete draft of the factual sections from research alone where possible.
## Step 3. Interview the user for gaps only
Each `steer-*` skill defines its own gap-topic list. Apply `./questioning-style.md` at that step.
## Step 4. Draft the document
Use the skill's reference template as the shape. Fill in all sections with gathered context. Replace every `[PLACEHOLDER]` with real content. Apply the skill's own writing rules. Apply strict STE per `./ste-writing.md` before writing the durable file. Build the DDD allowlist. Grep the STE dictionary. Run the self-lint.
## Step 5. Review with user
Show the draft. Apply `./questioning-style.md` and ask "Does this accurately reflect <the focus>?" — header `Review`:
- **Looks good — save it** → write to `docs/steering/<DOC>.md`
- **I have changes** → adjust first
Apply edits, then proceed.
## Step 6. Write and commit
```bash
mkdir -p docs/steering
# Write the final content to docs/steering/<DOC>.md
git add docs/steering/<DOC>.md
git commit -m "docs: add <subject> steering document"
```
Print the file path.
## Step 7. Offer wiki sync
Apply `./questioning-style.md` and ask "Would you like to sync this to the GitHub wiki?" — header `Wiki sync`:
- **Yes — push to wiki** → publish `<DOC>.md` as a wiki page
- **Not now** → skip wiki sync
If **yes**:
```bash
REPO=$(gh repo view --json nameWithOwner -q .nameWithOwner)
WIKI_DIR=$(mktemp -d -t wtf.wiki-sync-<slug>-XXXXXX)
git clone "https://github.com/$REPO.wiki.git" "$WIKI_DIR"
cp "docs/steering/<DOC>.md" "$WIKI_DIR/WTF-<Subject>.md"
(cd "$WIKI_DIR" && git add "WTF-<Subject>.md" && git commit -m "Sync: <subject>" && git push)
rm -rf "$WIKI_DIR"
```
Substitute `<slug>` with a lowercase word matching the doc (for example `tech`, `design`, `qa`, `vision`). Substitute `<Subject>` with the capitalized display name.
## Step 8. Offer to continue
Apply `./questioning-style.md` and ask "Continue with another steering doc?" — header `Next steering doc`:
- One option per remaining `steer-*` skill not yet created (excluding this skill) → run that skill
- **Stop here** → exit, no further action
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |