@kennethsolomon/shipkit 1.0.0

package/README.md

@@ -0,0 +1,321 @@
+# SHIPKIT
+
+A structured, quality-gated workflow system for Claude Code.
+
+Ship features with TDD, auto-detecting linters, security audits, and AI-powered code review — all wired into a single repeatable workflow.
+
+[![npm](https://img.shields.io/npm/v/@kennethsolomon%2Fshipkit)](https://www.npmjs.com/package/@kennethsolomon/shipkit)
+[![license](https://img.shields.io/badge/license-MIT-blue)](LICENSE)
+[![platform](https://img.shields.io/badge/platform-Mac%20%2F%20Linux%20%2F%20Windows-lightgrey)](#)
+
+```
+npx @kennethsolomon/shipkit
+```
+
+Works on Mac, Linux, and Windows.
+
+```
+~ $ npx @kennethsolomon/shipkit
+
+           ⚑
+           |
+          /|\
+         / | \
+        /  |  \
+       /   |   \
+      / ☠  |  ☠ \
+     /     |     \
+    /______|______\
+  ▄████████████████▄
+  █  ◉          ◉  █
+   ▀██████████████▀
+     ≋  ≋  ≋  ≋  ≋
+
+  ShipKit v1.0.0
+  A structured workflow toolkit for Claude Code.
+  by Kenneth Solomon
+
+  ✓ Installed commands/sk (15 commands)
+  ✓ Installed skills (19 skills)
+
+  Done! Run /sk:help to get started.
+
+~ $
+```
+
+---
+
+> "Stop winging it. Ship with a system."
+
+---
+
+## How It Works
+
+ShipKit installs a set of slash commands and skills into your Claude Code config (`~/.claude/`). Each command is a focused instruction set that Claude follows — no magic, just structured prompts that enforce quality gates.
+
+The workflow is linear: **Explore → Design → Plan → Branch → Test → Implement → Lint → Verify → Security → Review → Ship.**
+
+Every gate must pass before the next step. If lint fails, you fix it. If tests don't cover new code, you write them. Security issues block the PR. This isn't optional — it's the whole point.
+
+---
+
+## Installation
+
+```bash
+npx @kennethsolomon/shipkit
+```
+
+Or clone and install locally (symlinks — changes reflect immediately):
+
+```bash
+git clone https://github.com/kennethsolomon/shipkit.git
+cd shipkit
+./install.sh
+```
+
+### Update
+
+```bash
+npx @kennethsolomon/shipkit
+```
+
+Re-running always installs the latest version.
+
+### Uninstall
+
+```bash
+npx @kennethsolomon/shipkit --uninstall
+```
+
+---
+
+## Workflow
+
+### Feature Flow
+
+```
+Brainstorm → Plan → Branch → [Schema] → Write Tests → Implement → Commit
+  → Lint ✓ → Test ✓ → Security ✓ → Review ✓ → Update Task → Finish
+```
+
+| # | Command | Purpose |
+|---|---------|---------|
+| 1 | read `tasks/todo.md` | Pick the next task |
+| 2 | read `tasks/lessons.md` | Review past corrections |
+| 3 | `/sk:brainstorm` | Clarify requirements — no code |
+| 4 | `/sk:frontend-design` | UI design spec *(skip if backend-only)* |
+| 5 | `/sk:api-design` | API contracts *(skip if no new endpoints)* |
+| 6 | `/sk:accessibility` | WCAG 2.1 AA audit on design *(skip if no frontend)* |
+| 7 | `/sk:write-plan` | Write plan to `tasks/todo.md` |
+| 8 | `/sk:branch` | Create branch from current task |
+| 9 | `/sk:schema-migrate` | Schema change analysis *(skip if no DB changes)* |
+| 10 | `/sk:write-tests` | TDD red: write failing tests first |
+| 11 | `/sk:execute-plan` | TDD green: make tests pass |
+| 12 | `/sk:smart-commit` | Conventional commit |
+| 13 | **`/sk:lint`** | **GATE** — all linters must pass |
+| 14 | `/sk:smart-commit` | Auto-skip if already clean |
+| 15 | **`/sk:test`** | **GATE** — 100% coverage on new code |
+| 16 | `/sk:smart-commit` | Auto-skip if already clean |
+| 17 | **`/sk:security-check`** | **GATE** — 0 issues |
+| 18 | `/sk:smart-commit` | Auto-skip if already clean |
+| 19 | **`/sk:review`** | **GATE** — 0 issues including nitpicks |
+| 20 | `/sk:smart-commit` | Auto-skip if already clean |
+| 21 | `/sk:update-task` | Mark done, log completion |
+| 22 | `/sk:finish-feature` | Changelog + PR |
+| 23 | `/sk:release` | Version bump + tag *(optional)* |
+
+### Bug Fix Flow
+
+```
+Debug → Plan → Branch → Write Tests → Implement → Lint ✓ → Test ✓ → Security ✓ → Review ✓ → Finish
+```
+
+| # | Command | Purpose |
+|---|---------|---------|
+| 1 | `/sk:debug` | Root-cause analysis |
+| 2 | `/sk:write-plan` | Fix plan |
+| 3 | `/sk:branch` | Create branch |
+| 4 | `/sk:write-tests` | Reproduce the bug in a test |
+| 5 | `/sk:execute-plan` | Fix — make the test pass |
+| 6–9 | lint → test → security → review | Quality gates |
+| 10 | `/sk:finish-feature` | Changelog + PR |
+
+### Hotfix Flow
+
+For production issues that need to ship immediately. Skips brainstorm, design, and TDD. **Quality gates are non-negotiable even in a hotfix.**
+
+```
+Debug → Branch → Fix → Smoke Test → Lint ✓ → Test ✓ → Security ✓ → Review ✓ → Finish
+```
+
+| # | Command | Purpose |
+|---|---------|---------|
+| 1 | `/sk:debug` | Root-cause analysis — understand before touching code |
+| 2 | `/sk:branch` | Auto-named from the bug description |
+| 3 | implement directly | No write-tests phase — go straight to the fix |
+| 4 | run existing tests | Existing tests MUST still pass |
+| 5 | `/sk:smart-commit` | Commit the fix |
+| 6 | **`/sk:lint`** | **GATE** |
+| 7 | **`/sk:test`** | **GATE** |
+| 8 | **`/sk:security-check`** | **GATE** |
+| 9 | **`/sk:review`** | **GATE** |
+| 10 | `/sk:update-task` | Mark done |
+| 11 | `/sk:finish-feature` | Changelog + PR — mark as hotfix |
+
+After merging: add a regression test and a lesson to `tasks/lessons.md`.
+
+---
+
+## Commands
+
+### Planning & Design
+
+| Command | Description |
+|---------|-------------|
+| `/sk:brainstorm` | Explore requirements and design before writing any code |
+| `/sk:frontend-design` | Create UI design specs and mockups |
+| `/sk:api-design` | Design REST or GraphQL API contracts |
+| `/sk:accessibility` | WCAG 2.1 AA audit on design or existing frontend |
+| `/sk:write-plan` | Write a decision-complete plan to `tasks/todo.md` |
+| `/sk:plan` | Create or refresh task planning files |
+| `/sk:setup-claude` | Bootstrap project scaffolding (CLAUDE.md + tasks/) |
+| `/sk:setup-optimizer` | Enrich CLAUDE.md by scanning the codebase |
+
+### Development
+
+| Command | Description |
+|---------|-------------|
+| `/sk:branch` | Create a feature branch from the current task |
+| `/sk:schema-migrate` | Analyze pending schema changes (Prisma, Drizzle, Eloquent, SQLAlchemy, ActiveRecord) |
+| `/sk:write-tests` | TDD: write failing tests before implementation |
+| `/sk:execute-plan` | Implement the plan in small batches |
+| `/sk:debug` | Structured bug investigation: reproduce → isolate → fix |
+| `/sk:hotfix` | Emergency fix workflow — skips design and TDD |
+
+### Quality Gates
+
+| Command | Description |
+|---------|-------------|
+| `/sk:lint` | Auto-detect and run all linters (Pint, ESLint, PHPStan, Prettier…) |
+| `/sk:test` | Auto-detect and run all test suites, verify 100% coverage on new code |
+| `/sk:security-check` | OWASP security audit across changed code |
+| `/sk:perf` | Performance audit: bundle size, N+1 queries, Core Web Vitals |
+| `/sk:review` | Rigorous self-review across 7 dimensions |
+
+### Shipping
+
+| Command | Description |
+|---------|-------------|
+| `/sk:smart-commit` | Generate conventional commit messages with approval |
+| `/sk:update-task` | Mark current task done, log completion |
+| `/sk:finish-feature` | Write changelog entry + create PR |
+| `/sk:release` | Version bump + CHANGELOG + git tag + push |
+| `/sk:features` | Sync docs/features/ specs with the codebase |
+
+### Laravel
+
+| Command | Description |
+|---------|-------------|
+| `/sk:laravel-new` | Scaffold a fresh Laravel app with production-ready conventions |
+| `/sk:laravel-init` | Configure an existing Laravel project |
+
+### Configuration
+
+| Command | Description |
+|---------|-------------|
+| `/sk:config` | View and edit project config (`.shipkit/config.json`) |
+| `/sk:set-profile` | Switch model routing profile for this project |
+
+### Meta
+
+| Command | Description |
+|---------|-------------|
+| `/sk:help` | Show all commands and workflow overview |
+| `/sk:status` | Show workflow and task status at a glance |
+| `/sk:skill-creator` | Create or improve ShipKit skills |
+
+---
+
+## Model Routing Profiles
+
+ShipKit routes each skill to the right model automatically based on your project profile.
+Set it once with `/sk:set-profile <name>` — config is saved to `.shipkit/config.json`.
+
+| Profile | Philosophy | Best for |
+|---------|-----------|---------|
+| `full-sail` | Opus on everything that matters | High-stakes work, client projects |
+| `quality` | Opus for planning + review, Sonnet for implementation | Most professional projects |
+| `balanced` | Sonnet across the board *(default)* | Day-to-day development |
+| `budget` | Haiku where possible, Sonnet for gates | Side projects, prototyping |
+
+| Skill group | full-sail | quality | balanced | budget |
+|-------------|-----------|---------|----------|--------|
+| brainstorm, write-plan, debug, execute-plan, review | opus | opus | sonnet | sonnet |
+| write-tests, frontend-design, api-design, security-check | opus | sonnet | sonnet | sonnet |
+| perf, schema-migrate, accessibility | opus | sonnet | sonnet | haiku |
+| lint, test | sonnet | sonnet | haiku | haiku |
+| smart-commit, branch, update-task | haiku | haiku | haiku | haiku |
+
+`opus` = inherit (uses your current session model).
+
+### Other config settings
+
+```json
+{
+  "profile": "balanced",
+  "auto_commit": true,
+  "skip_gates": [],
+  "coverage_threshold": 100,
+  "branch_pattern": "feature/{slug}",
+  "model_overrides": { "sk:review": "opus" }
+}
+```
+
+| Setting | Default | Description |
+|---------|---------|-------------|
+| `profile` | `balanced` | Model routing profile |
+| `auto_commit` | `true` | Auto-commit after each gate passes |
+| `skip_gates` | `[]` | Gates to skip — e.g. `["perf","accessibility"]` for backend-only projects |
+| `coverage_threshold` | `100` | Minimum test coverage % on new code |
+| `branch_pattern` | `feature/{slug}` | Branch naming convention |
+| `model_overrides` | `{}` | Per-skill model overrides that take precedence over profile |
+
+---
+
+## Stack Support
+
+ShipKit auto-detects your stack — no configuration needed.
+
+| Area | Supported |
+|------|-----------|
+| **Linters** | Pint, ESLint, PHPStan, Rector, Prettier, Biome, Stylelint |
+| **Test runners** | Pest, PHPUnit, Jest, Vitest, Playwright |
+| **Schema / ORM** | Prisma, Drizzle, Eloquent, SQLAlchemy + Alembic, ActiveRecord |
+| **Frameworks** | Laravel, Next.js, Nuxt, React, Vue, Node.js |
+| **Release** | npm, Composer, iOS (App Store), Android (Play Store) |
+
+---
+
+## Project Setup
+
+To wire ShipKit into a new project:
+
+```
+/sk:setup-claude
+```
+
+This creates `tasks/todo.md`, `tasks/lessons.md`, and a project-specific `CLAUDE.md` with the full workflow baked in.
+
+---
+
+## Why ShipKit
+
+Claude Code is powerful but context degrades as the window fills. Unstructured sessions lead to skipped tests, no lint, missing security review, and PRs that broke things in ways nobody caught.
+
+ShipKit fixes that with a repeatable system: every feature goes through the same gates in the same order. Quality isn't optional — it's structural.
+
+---
+
+## License
+
+MIT — by [Kenneth Solomon](https://github.com/kennethsolomon)

package/bin/shipkit.js

@@ -0,0 +1,146 @@
+#!/usr/bin/env node
+'use strict';
+
+const fs   = require('fs');
+const path = require('path');
+const os   = require('os');
+
+const pkg = require('../package.json');
+
+// ── ANSI ────────────────────────────────────────────────────────────────────
+const cyan   = '\x1b[36m';
+const green  = '\x1b[32m';
+const yellow = '\x1b[33m';
+const bold   = '\x1b[1m';
+const dim    = '\x1b[2m';
+const reset  = '\x1b[0m';
+
+// ── Pirate ship ASCII art ────────────────────────────────────────────────────
+const banner = `
+${cyan}           ⚑${reset}
+${cyan}           |${reset}
+${cyan}          /|\\${reset}
+${cyan}         / | \\${reset}
+${cyan}        /  |  \\${reset}
+${cyan}       /   |   \\${reset}
+${cyan}      / ☠  |  ☠ \\${reset}
+${cyan}     /     |     \\${reset}
+${cyan}    /______|______\\${reset}
+${cyan}  ▄████████████████▄${reset}
+${cyan}  █  ◉          ◉  █${reset}
+${cyan}   ▀██████████████▀${reset}
+${cyan}     ≋  ≋  ≋  ≋  ≋${reset}
+
+  ${bold}ShipKit${reset} ${dim}v${pkg.version}${reset}
+  A structured workflow toolkit for Claude Code.
+  by Kenneth Solomon
+
+`;
+
+// ── Helpers ──────────────────────────────────────────────────────────────────
+function getClaudeDir() {
+  if (process.platform === 'win32') {
+    return path.join(process.env.APPDATA || os.homedir(), 'Claude');
+  }
+  return path.join(os.homedir(), '.claude');
+}
+
+function copyDir(src, dest) {
+  fs.mkdirSync(dest, { recursive: true });
+  for (const entry of fs.readdirSync(src, { withFileTypes: true })) {
+    const srcPath  = path.join(src, entry.name);
+    const destPath = path.join(dest, entry.name);
+    if (entry.isDirectory()) {
+      copyDir(srcPath, destPath);
+    } else {
+      fs.copyFileSync(srcPath, destPath);
+    }
+  }
+}
+
+function countFiles(dir, ext) {
+  if (!fs.existsSync(dir)) return 0;
+  return fs.readdirSync(dir).filter(f => f.endsWith(ext)).length;
+}
+
+// ── Uninstall ────────────────────────────────────────────────────────────────
+function uninstall() {
+  process.stdout.write(banner);
+  const claudeDir   = getClaudeDir();
+  const commandsDest = path.join(claudeDir, 'commands', 'sk');
+  const skillsDest   = path.join(claudeDir, 'skills');
+
+  if (fs.existsSync(commandsDest)) {
+    fs.rmSync(commandsDest, { recursive: true, force: true });
+    console.log(`  ${green}✓${reset} Removed commands/sk`);
+  }
+
+  // Remove only sk: prefixed skill dirs
+  if (fs.existsSync(skillsDest)) {
+    const removed = fs.readdirSync(skillsDest, { withFileTypes: true })
+      .filter(e => e.isDirectory() && e.name.startsWith('sk:'))
+      .map(e => {
+        fs.rmSync(path.join(skillsDest, e.name), { recursive: true, force: true });
+        return e.name;
+      });
+    console.log(`  ${green}✓${reset} Removed ${removed.length} skills`);
+  }
+
+  console.log(`\n  ${green}Done!${reset} ShipKit uninstalled.\n`);
+}
+
+// ── Install ──────────────────────────────────────────────────────────────────
+function install() {
+  process.stdout.write(banner);
+
+  const claudeDir  = getClaudeDir();
+  const pkgDir     = path.join(__dirname, '..');
+
+  // Check Claude dir exists
+  if (!fs.existsSync(claudeDir)) {
+    fs.mkdirSync(claudeDir, { recursive: true });
+  }
+
+  // Install commands/sk/
+  const commandsSrc  = path.join(pkgDir, 'commands', 'sk');
+  const commandsDest = path.join(claudeDir, 'commands', 'sk');
+
+  if (fs.existsSync(commandsSrc)) {
+    copyDir(commandsSrc, commandsDest);
+    const count = countFiles(commandsSrc, '.md');
+    console.log(`  ${green}✓${reset} Installed commands/sk ${dim}(${count} commands)${reset}`);
+  } else {
+    console.log(`  ${yellow}!${reset} commands/sk not found — skipping`);
+  }
+
+  // Install skills/sk:*/
+  const skillsSrc  = path.join(pkgDir, 'skills');
+  const skillsDest = path.join(claudeDir, 'skills');
+  let skillCount = 0;
+
+  if (fs.existsSync(skillsSrc)) {
+    const skillDirs = fs.readdirSync(skillsSrc, { withFileTypes: true })
+      .filter(e => e.isDirectory());
+
+    for (const entry of skillDirs) {
+      const src  = path.join(skillsSrc, entry.name);
+      const dest = path.join(skillsDest, entry.name);
+      copyDir(src, dest);
+      skillCount++;
+    }
+    console.log(`  ${green}✓${reset} Installed skills ${dim}(${skillCount} skills)${reset}`);
+  } else {
+    console.log(`  ${yellow}!${reset} skills/ not found — skipping`);
+  }
+
+  console.log(`\n  ${green}Done!${reset} Run ${cyan}/sk:help${reset} to get started.\n`);
+}
+
+// ── Entry point ──────────────────────────────────────────────────────────────
+const args = process.argv.slice(2);
+
+if (args.includes('--uninstall') || args.includes('-u')) {
+  uninstall();
+} else {
+  install();
+}

package/commands/sk/brainstorm.md

@@ -0,0 +1,63 @@
+---
+description: "Start with design questions before writing code."
+---
+
+<!-- Generated by /sk:setup-claude -->
+
+# /sk:brainstorm
+
+Explore design and clarify requirements **before** any code is written.
+
+## Hard Rules
+
+- **DO NOT write, edit, or generate any code.** No files, no snippets, no pseudo-implementations.
+- **DO NOT create a plan.** That is `/sk:write-plan`'s job.
+- **DO NOT run build/sk:test/sk:lint commands.** You are in design mode only.
+- You may **read** existing code to understand the current state, but nothing more.
+
+## Steps
+
+1. **Read context files first:**
+   - If `tasks/findings.md` exists and has content, read **only sections relevant to the
+     current phase/task** — skip decisions from completed or unrelated phases. Summarize
+     prior decisions and ask: extend, revise, or start fresh?
+   - If `tasks/lessons.md` exists, read it in full. Apply every active lesson as a design
+     constraint throughout this brainstorm — treat each prevention rule as a hard constraint
+     when proposing approaches.
+   - If `tasks/security-findings.md` exists and has audit results, skim for recurring
+     patterns — factor security constraints into proposed approaches (e.g., if prior audits
+     flagged missing input validation, ensure new designs account for it).
+
+2. **Understand the request** — Ask clarifying questions about what the user wants. Do not assume. Cover:
+   - What problem are we solving?
+   - Who is it for?
+   - What are the expected inputs/outputs or behaviors?
+   - Are there constraints (performance, compatibility, existing patterns)?
+
+3. **Parallel codebase research** — Launch 3 Explore agents **in parallel** (read-only) before proposing approaches:
+   - **Agent 1 — Affected areas:** Explore models, services, controllers, and files directly related to the task. Map the relevant code surface.
+   - **Agent 2 — Existing patterns:** Explore how similar features were built in this project. Identify conventions, shared abstractions, and naming patterns to follow.
+   - **Agent 3 — Tests & edge cases:** Explore existing test coverage for the affected area. Identify gaps, edge cases, and validation patterns already in use.
+
+   Wait for all three agents to return, then synthesize their findings into a brief summary before moving to step 4.
+
+4. **Propose approaches** — Present 2–3 design options with trade-offs:
+   - Approach name + 1-sentence summary
+   - Pros and cons
+   - Complexity estimate (small / medium / large)
+   - Which existing patterns it follows or breaks
+
+5. **Get alignment** — Ask the user which approach they prefer (or if they want a hybrid). Do not proceed without explicit approval on the direction.
+
+6. **Record findings** — Write discoveries and the agreed-upon direction to `tasks/findings.md`:
+   - Problem statement
+   - Key decisions made
+   - Chosen approach + rationale
+   - Open questions (if any remain)
+
+## When Done
+
+Tell the user:
+> "Brainstorming complete. Findings saved to `tasks/findings.md`. Run `/sk:write-plan` to create the implementation plan."
+
+**Do not proceed to planning or implementation yourself.** The user must explicitly invoke the next step.

package/commands/sk/branch.md

@@ -0,0 +1,35 @@
+---
+description: "Create a feature branch from the current task in tasks/todo.md."
+---
+
+# /sk:branch
+
+Create a feature branch for the current task.
+
+## Steps
+
+### 1. Read Current Task
+- Read `tasks/todo.md` and find the next incomplete task (first unchecked checkbox)
+- Extract a short name from the task description
+
+### 2. Check Prerequisites
+- Ensure working directory is clean (`git status`). If dirty, warn the user and stop.
+- Ensure you're on the main branch. If not, ask the user to confirm branching from the current branch.
+
+### 3. Pull Latest
+```bash
+git pull origin main
+```
+
+### 4. Create Branch
+- Generate branch name: `feature/<short-task-name>` (kebab-case, max 50 chars)
+  - Example task: "Add server metrics dashboard" → `feature/server-metrics-dashboard`
+  - Example task: "Fix duplicate episode notifications" → `fix/duplicate-episode-notifications`
+- Use `fix/` prefix for bug fixes, `feature/` for new features, `refactor/` for refactoring
+```bash
+git checkout -b <branch-name>
+```
+
+### 5. Confirm
+- Show the user the branch name created
+- Show which task this branch is for

package/commands/sk/config.md

@@ -0,0 +1,96 @@
+---
+description: "View and edit ShipKit project configuration (.shipkit/sk:config.json)."
+---
+
+# /sk:config
+
+View and manage ShipKit configuration for this project.
+
+## Config File
+
+ShipKit stores per-project config in `.shipkit/sk:config.json` at the project root.
+If the file does not exist, display defaults and offer to create it.
+
+## Steps
+
+### 1 — Read current config
+
+Check if `.shipkit/sk:config.json` exists. If yes, read and parse it.
+If not, use defaults:
+
+```json
+{
+  "profile": "balanced",
+  "auto_commit": true,
+  "skip_gates": [],
+  "coverage_threshold": 100,
+  "branch_pattern": "feature/{slug}",
+  "model_overrides": {}
+}
+```
+
+### 2 — Display current config
+
+Show a formatted table explaining every setting and its current value:
+
+| Setting | Value | Description |
+|---------|-------|-------------|
+| `profile` | `<value>` | Model routing profile (full-sail / quality / balanced / budget) |
+| `auto_commit` | `<value>` | Auto-run `/sk:smart-commit` after each gate passes |
+| `skip_gates` | `<value>` | Gates to skip (e.g. `["perf","accessibility"]` for backend-only projects) |
+| `coverage_threshold` | `<value>` | Minimum test coverage % on new code (default: 100) |
+| `branch_pattern` | `<value>` | Branch naming pattern (`feature/{slug}`, `feat/{slug}`, etc.) |
+| `model_overrides` | `<value>` | Per-skill model overrides e.g. `{ "sk:review": "opus" }` |
+
+Then show the model table for the **current profile**:
+
+#### Model assignments — `<profile>` profile
+
+| Skill | Model |
+|-------|-------|
+| brainstorm, write-plan, debug, execute-plan, review | `<model>` |
+| write-tests, frontend-design, api-design, security-check | `<model>` |
+| perf, schema-migrate, accessibility | `<model>` |
+| lint, test | `<model>` |
+| smart-commit, branch, update-task | `haiku` |
+
+Use this model table to fill in values:
+
+| Skill | full-sail | quality | balanced | budget |
+|-------|-----------|---------|----------|--------|
+| brainstorm, write-plan, debug, execute-plan, review | opus | opus | sonnet | sonnet |
+| write-tests, frontend-design, api-design, security-check | opus | sonnet | sonnet | sonnet |
+| perf, schema-migrate, accessibility | opus | sonnet | sonnet | haiku |
+| lint, test | sonnet | sonnet | haiku | haiku |
+| smart-commit, branch, update-task | haiku | haiku | haiku | haiku |
+
+Note: `opus` resolves to `inherit` — Claude uses whatever model the current session is running on.
+
+### 3 — Offer to edit
+
+Ask: "Which setting would you like to change? (or press Enter to exit)"
+
+If the user specifies a setting:
+
+- `profile` → show `/sk:set-profile` options and apply
+- `auto_commit` → toggle true/false
+- `skip_gates` → show available gates (`lint`, `test`, `security-check`, `perf`, `accessibility`, `review`) and let user pick
+- `coverage_threshold` → accept a number 0-100
+- `branch_pattern` → accept a pattern string (show examples: `feature/{slug}`, `feat/{slug}`, `{slug}`)
+- `model_overrides` → accept JSON or guided per-skill input
+
+### 4 — Write config
+
+If any setting was changed:
+1. Create `.shipkit/` directory if it does not exist
+2. Write updated config to `.shipkit/sk:config.json`
+3. Add `.shipkit/sk:config.json` to `.gitignore` if not already present
+4. Confirm: "Config saved to `.shipkit/sk:config.json`"
+
+If no changes, exit cleanly.
+
+## Notes
+
+- Config is **per-project** — each repo has its own `.shipkit/sk:config.json`
+- `model_overrides` takes precedence over `profile` for individual skills
+- Gitignore: `.shipkit/sk:config.json` is personal preference — don't commit it unless the team agrees on a shared profile