Setup Repo

# Code

## Code Structure
- **<component>:** `<path>` (<brief description>)

## Branch Naming
- **Bug-fix branches:** <pattern, e.g. `<name>/bugs/<ticket-id>`>
- **Feature branches:** <pattern, e.g. `<name>/<feature-description>`>
- **Example:** <e.g. `jane/bugs/PROJ-42`, `fix/login-crash`>

## Build
- **Full build:** `<command>`

## Run
- **Dev mode:** `<command>`

## Pre-commit Verification
- **Format:** `<command>`
- **Check (lint + types):** `<command>`
- **Unit tests:** `<command>`

## Publishing Changes
- **Push command:** `<command, e.g. git push -u origin <branch>>`
- **Create MR/PR (base command):** `<assembled command WITHOUT --fill, e.g. glab mr create --target-branch main --remove-source-branch>` <!-- autonomous skills append --title / --description (glab) or --title / --body (gh) at runtime -->
- **Auto-publish allowed:** `<yes | no>` <!-- read by autonomous skills; if "no", they stop after committing and let the user publish -->

### Merge defaults
- **Delete source branch on merge:** `<yes | no>`
- **Squash on merge:** `<yes | no>`
- **Auto-merge when CI passes:** `<yes | no>`
- **Open as draft:** `<yes | no>`

### Conventions (optional)
- <required labels, reviewers, title format, body template>

## Proof of Work

Every MR/PR opened by an autonomous skill must include evidence that the
bug existed and is now fixed. The MR/PR body walks a reviewer through:

1. **Original bug** — exact description (and ticket link, if any).
2. **Reproduction** — repro steps, plus before-evidence (screenshots / logs /
   error trace) showing the buggy behavior.
3. **Hypothesis** — what code path was responsible and why.
4. **Fix** — what changed and why it addresses the cause.
5. **Proof the fix works** — after-evidence (screenshots / test output /
   logs) showing the correct behavior, plus a link or hash for the
   failing-test commit.

### Optional sections (only when applicable)
- **`## Deferred follow-ups`** — list any adjacent work the agent
  consciously left out of this MR. If the testing config has a
  "How to file new tickets" entry point, each item MUST also have a
  tracker ticket URL next to it. Omit the section entirely if no
  work was deferred.
- **`## Review notes`** — populated by the fix-bug self-review pass
  with any code-review findings the agent declined to act on.
  Required after every autonomous fix-bug run: if the agent acted on
  every finding (or found none), write `_(no outstanding review
  notes)_` so a reviewer knows the review pass ran.

### Evidence tooling
- **UI-visible bugs:** <e.g. capture before/after with `/auto-qa-changes`>
- **Non-UI bugs:** <e.g. paste the failing-then-passing test output from `/run-integration-test`>
- **Other artifacts (optional):** <e.g. logs, benchmark deltas, curl transcripts>

### Required vs optional
- **Required for every MR/PR:** <yes | no, with conditions> <!-- e.g. "yes for any UI-visible change" -->

## Dependencies (optional)
- **Install:** `<command>`

## Environment Setup (optional)
- <any required env vars, config files, or services>

## Deployment (optional)
- <how to deploy or release>

# Docs

Locations and operational config for the sculptor-workflow skills.
Document structure (spec sections, architecture sections, mock
conventions, plan layout) is baked into the skills — not configured
here.

## Spec Location
- **Path pattern:** <from the user's answer, e.g. `specs/<slug>/spec.md`>

## UI Reference

<!--
  Optional: tell the `/sculptor-workflow:mock` skill how to match your app's
  visual style. Examples:
    - "Read frontend/src/components/ for the component library."
    - "See docs/design-system.md for tokens and spacing."
    - "Use the Storybook at localhost:6006."
    - "Standalone project — no existing app to match."
  If empty, the mock skill will scan the repo for UI code on its own.
  If there is neither a UI Reference nor discoverable frontend code, the
  mock skill will ask for direction before generating.
-->

## Code Review

The configured skill below is invoked for code-review passes by:
- `/sculptor-workflow:review` at the end of the full feature workflow.
- `/sculptor-workflow:fix-bug`'s self-review phase (Phase 4 interactive
  / Phase A4.5 autonomous), so a fix isn't considered done until its
  diff has been reviewed.

<!--
  Set this to your repo's review skill, e.g. /code-review-checklist.
  If left empty, both /review and fix-bug's self-review phase skip
  the code-review pass.
-->

Skill: <skill-name>

# Testing

## Test Strategy
<when to write e2e vs unit tests, and when to skip tests entirely>

## Test Framework
- **Framework:** <name>
- **Run a test file:** `<command>`
- **Run a single test:** `<command>`
- **Run e2e tests:** <command or skill — omit this line if e2e tests run with the same command as unit tests>
- **Test location:** <where test files go>
- **Conventions:** <naming, fixtures, patterns, or "see <file> as reference">

## Bug Tracking (optional)
- **System:** <name>
- **Ticket ID format:** <pattern, e.g. SCU-123, PROJ-42>
- **How to fetch ticket context:** <instructions or skill name>
- **When to fetch:** <e.g. "whenever input matches ticket format", "always required", "only if user provides a ticket ID">
- **How to file new tickets:** <instructions or skill name + entry point, e.g. "use the `/linear` skill's `create-ticket` entry point"> <!-- read by /fix-bug autonomous mode to file deferred-work follow-ups -->
- **How to comment on tickets:** <instructions or skill name + entry point, e.g. "use the `/linear` skill's `comment` entry point"> <!-- read by /fix-bug autonomous mode to leave evidence on tickets the agent declines to fix -->
- **How to change ticket state:** <instructions or skill name + entry point, e.g. "use the `/linear` skill's `set-state` entry point"> <!-- read by /fix-bug autonomous mode to move tickets to a "needs info" state -->
- **Needs-info state name:** <workflow-state name, e.g. `Triage`, `Needs Info`, `Awaiting Reporter`> <!-- the state /fix-bug autonomous mode moves a ticket to when it concludes the bug is unreproducible from the description -->


## Manual Testing (optional)
- **How to test:** <instructions or skill name>
- **Notes:** <any setup steps, URLs, or caveats>

## Visual Verification (optional)
- **How to verify:** <instructions or skill name>

## Test Debugging (optional)
- **How to debug:** <instructions or skill name>

## Test Writing (optional)
- **How to write tests:** <instructions or skill name>
- **Test types and locations:** <if there are multiple categories>

git add .sculptor/code.md .sculptor/testing.md .sculptor/docs.md
git commit -m "Add Sculptor repo config"