minimal-vibe-coding-kit 0.4.0 → 0.4.2

package/.grok/skills/visual-design-loop/SKILL.md

@@ -0,0 +1,103 @@
+---
+name: visual-design-loop
+description: Run an iterative visual design improvement loop for UI/product polish tasks. Use when a request involves screenshots, rendering, visual review, frontend polish, layout/typography/color refinement, design QA, or a Claude/Codex loop goal that touches a visible product surface.
+---
+
+# Visual Design Loop
+
+Use this skill to improve a visible UI through controlled render-review-fix loops. Keep the brief and existing project conventions as the source of truth; do not add new product requirements just because a screen could be enhanced.
+
+## Contract
+
+Before editing, resolve and state:
+
+- Goal.
+- Target surface: route, page, component, flow, or supplied screenshot.
+- Source of truth: current brief plus repo instructions, design system, nearby UI, and `backbone.yml` when present.
+- Render method: local URL, app command, Storybook story, preview command, static file, or supplied screenshot.
+- Screenshot method: browser screenshot, Playwright, in-app browser, provided image, or equivalent.
+- Editable paths.
+- Protected paths.
+- Budget: default 3 loops.
+- Timeout: default 10 minutes per loop.
+- Log path: `/tmp/design-{project_slug}.md`.
+
+If render or screenshot capture is missing and cannot be inferred safely, ask for the missing command, URL, or image before editing.
+
+## Baseline
+
+1. Render the target surface.
+2. Capture or inspect the baseline screenshot.
+3. Review visible issues only; do not invent findings.
+4. Create or update `/tmp/design-{project_slug}.md`.
+5. Log the baseline screenshot, visual score, main issues, and any render constraints.
+
+Derive `project_slug` from `backbone.yml` `project.name` when available; otherwise use the repository directory name.
+
+## Visual rubric
+
+Score screenshots from 1 to 5 using:
+
+- Visual hierarchy: primary content and action are obvious.
+- Layout and spacing: alignment, rhythm, density, whitespace.
+- Typography: scale, contrast, readability, consistency.
+- Color and contrast: accessible, intentional, not noisy.
+- Component consistency: matches nearby UI and design system.
+- Responsiveness: works at relevant viewport sizes.
+- Interaction states: hover, focus, disabled, empty, loading, and error states when relevant.
+- Product fit: improves the brief without adding unnecessary features.
+
+## Loop
+
+For each loop:
+
+1. Review the latest screenshot.
+2. Identify the highest-impact visual issue.
+3. Form one small hypothesis.
+4. Apply one targeted fix inside editable paths only.
+5. Render again.
+6. Capture the after screenshot.
+7. Compare before and after.
+8. Keep the change only if quality improves, or if quality is equal with simpler and safer implementation.
+9. Revert only your own loop changes if the result is worse.
+10. Run the repo validation command when code changed.
+11. Append the loop result to `/tmp/design-{project_slug}.md`.
+
+Each loop entry must include:
+
+- Loop number.
+- Screenshot reviewed.
+- Issue found.
+- Hypothesis.
+- Fix applied.
+- Before/after judgment.
+- Rubric score before.
+- Rubric score after.
+- Validation result.
+- Remaining concerns.
+- Stop/continue decision.
+
+## Stop criteria
+
+Stop when:
+
+- The design is polished enough for the brief.
+- The latest loop gives no meaningful improvement.
+- Further changes would be subjective or overworked.
+- Budget is reached.
+- Validation fails for a non-trivial reason.
+- A product or design decision is required from the user.
+- Render or screenshot tooling is unavailable.
+
+## Final report
+
+Report:
+
+- Baseline score.
+- Final score.
+- Screenshots reviewed.
+- Changes kept.
+- Changes discarded, if any.
+- Validation result.
+- Remaining risks or design decisions.
+- Log path.

package/.grok/skills/visual-design-loop/agents/openai.yaml

@@ -0,0 +1,4 @@
+interface:
+  display_name: "Visual Design Loop"
+  short_description: "Iterative screenshot-based UI polish loop"
+  default_prompt: "Use $visual-design-loop to render, screenshot, review, and improve this UI until it is polished but not overworked."

package/.vibekit/docs/BACKBONE_REFERENCE.md

@@ -10,7 +10,7 @@
 - `conventions`: user-reviewed project rules inferred from existing code.
 - `commands`: install, test, lint, typecheck, build, validate.
 - `policy`: branch, commit, editable paths, protected paths.
-- `agent_surfaces`: Claude, Cursor, Codex, shared skills, shared commands.
+- `agent_surfaces`: Claude, Cursor, Codex, Grok, shared skills, shared commands.
 - `automation`: autoresearch, daily enhance, finalize (graduation cleanup), security.
 
 ## Init status

package/.vibekit/docs/INSTALL.md

@@ -1,5 +1,17 @@
 # Install guide
 
+## Install from npm
+
+```bash
+# one-shot (recommended) — adds nothing to the project's dependencies:
+npx --yes minimal-vibe-coding-kit@latest install /path/to/project
+
+# or, after `npm i -D minimal-vibe-coding-kit` inside the project:
+npx mvck install .
+```
+
+`npm i` alone only places the kit in `node_modules/` and nothing is active yet; `mvck install` (alias: `vibe-kit`) is the step that copies the kit files into the repo root.
+
 ## Local install
 
 ```bash
@@ -14,9 +26,10 @@ node .vibekit/scripts/mvck.mjs install /path/to/project --profile all
 
 Profiles:
 
-- `claude`: `CLAUDE.md`, `.claude/`, Claude skills, agents, commands, rules.
-- `cursor`: `.cursor/rules`, `.cursor/commands`, Cursor skill entrypoints.
-- `codex`: `AGENTS.md`, `.agents/skills`, `.codex-plugin`, `.codex` examples.
+- `claude`: `CLAUDE.md`, `.claude/`, Claude skills, agents, commands, rules, deny-list settings.
+- `cursor`: `.cursor/rules`, `.cursor/commands`, `.cursor/cli.json` CLI permissions, Cursor skill entrypoints.
+- `codex`: `AGENTS.md`, `.agents/skills`, `.codex/rules` execution-policy rules, `.codex-plugin`, `.codex` examples.
+- `grok`: `AGENTS.md`, `.grok/rules`, `.grok/skills`, `.grok/config.toml` permission rules, user config example.
 - `all`: every profile.
 
 ## Safe behavior
@@ -47,7 +60,7 @@ node /path/to/kit/.vibekit/scripts/mvck.mjs update /path/to/project
 
 The updater:
 
-- refreshes kit-owned surfaces (`.vibekit/skills/`, `.vibekit/commands/`, `.vibekit/scripts/`, `.vibekit/docs/`, and the `.claude/`, `.cursor/`, `.agents/`, `.codex*` mirrors) and adds any new kit skills;
+- refreshes kit-owned surfaces (`.vibekit/skills/`, `.vibekit/commands/`, `.vibekit/scripts/`, `.vibekit/docs/`, and the `.claude/`, `.cursor/`, `.agents/`, `.grok/`, `.codex*` mirrors) and adds any new kit skills;
 - never overwrites `backbone.yml`, `CLAUDE.md`, `AGENTS.md` content outside the managed block, or `settings.json` files — those are seeded only if missing;
 - backs up every replaced kit file to `.vibekit/update-backup/<timestamp>/` (disable with `--no-backup`);
 - never deletes files you added, and skips re-seeding one-time files after `mvck finalize`;
@@ -102,7 +115,7 @@ node .vibekit/scripts/mvck.mjs doctor . --write-report
 
 ## Native reasoning skills
 
-The installer includes three flexible custom reasoning skills across Claude, Codex, and Cursor. These install as full skill folders, including examples and references for progressive disclosure:
+The installer includes three flexible custom reasoning skills across Claude, Codex, Cursor, and Grok. These install as full skill folders, including examples and references for progressive disclosure:
 
 - `clearthought`: clarify ambiguous tasks and choose a reasoning mode.
 - `sequential-thinking`: split complex work into ordered implementation steps.
@@ -110,11 +123,12 @@ The installer includes three flexible custom reasoning skills across Claude, Cod
 
 ## Visual design loop skill
 
-The installer also includes `visual-design-loop` for Claude and Codex surfaces. Use it when a loop goal touches UI polish, screenshots, rendering, visual QA, or visible frontend behavior.
+The installer also includes `visual-design-loop` for Claude, Codex, and Grok surfaces. Use it when a loop goal touches UI polish, screenshots, rendering, visual QA, or visible frontend behavior.
 
 ## User-invoked utility skills
 
-Two user-invoked skills install across Claude, Codex, and Cursor surfaces:
+Three user-invoked skills install across Claude, Codex, Cursor, and Grok surfaces:
 
 - `memento`: write a `MEMENTO.md` working note before closing a multi-day task (`/memento`), then resume from it in the next session (`/memento resume`).
 - `coding-level`: set the explanation register from 0 (ELI5) to 5 (expert peer) with `/coding-level N`; stays active until reinvoked.
+- `prompt-sharpener`: sharpen a rough prompt into a precise one with `/prompt-sharpener <rough prompt>`, then execute the sharpened version immediately in the same turn.

package/.vibekit/docs/SECURITY_MODEL.md

@@ -5,7 +5,7 @@
 Review these before merging changes:
 
 - `CLAUDE.md`, `AGENTS.md`, `.github/copilot-instructions.md`
-- `.claude/**`, `.cursor/**`, `.agents/**`, `.codex-plugin/**`
+- `.claude/**`, `.cursor/**`, `.agents/**`, `.grok/**`, `.codex-plugin/**`
 - `.vibekit/skills/**`, `.vibekit/commands/**`, `.vibekit/scripts/**`, `hooks/**`
 - `.mcp.json`, `mcp.json`, `mcp-configs/**`
 - installer scripts and CI workflows

package/.vibekit/docs/TOOLING_GUIDE.md

@@ -12,11 +12,16 @@ Use `.cursor/rules/*.mdc`. The kit creates always-on core rules plus scoped rule
 
 Use `AGENTS.md` for repo instructions and `.agents/skills/*/SKILL.md` for skills. The kit also includes `.codex-plugin/plugin.json` so the shared skills can be packaged as a plugin.
 
+## Grok
+
+Grok Build reads `AGENTS.md` natively plus every `*.md` rule in `.grok/rules/`, and repository skills from `.grok/skills/*/SKILL.md` (user-invocable skills appear as `/<skill-name>` slash commands). Permission deny rules are user-level only: see `.grok/config.example.toml` for the `~/.grok/config.toml` setup.
+
 ## Shared skill strategy
 
 Canonical skill bodies live in `.vibekit/skills/<skill-name>/`. Harness-specific folders contain small shims:
 
 - `.claude/skills/<skill-name>/SKILL.md`
 - `.agents/skills/<skill-name>/SKILL.md`
+- `.grok/skills/<skill-name>/SKILL.md`
 
 This reduces drift while still giving each tool the path it expects.

package/.vibekit/docs/templates/CONTEXT_TEMPLATE.md

@@ -1,7 +1,7 @@
 # Project context
 
 > Scaffolded by Minimal Vibe Coding Kit. A short, shared glossary and map so any
-> agent (Claude, Cursor, Codex) navigates this project with less back-and-forth.
+> agent (Claude, Cursor, Codex, Grok) navigates this project with less back-and-forth.
 > Keep it current and terse — definitions, not prose.
 
 ## Domain glossary

package/.vibekit/init/CLAUDE-template.md

@@ -16,11 +16,14 @@ This file is intentionally small. Shared rules live in `AGENTS.md`, project fact
 - `/autoresearch-coding` - run a metric-driven improvement loop.
 - `/security-scan` - run AgentShield-style review.
 - `/daily-enhance` - propose rule, skill, and workflow improvements.
+- `/coding-level N` - set explanation depth 0-5; the project default lives in `backbone.yml` `conventions.custom_rules`.
+- `/prompt-sharpener <rough prompt>` - sharpen a rough prompt into a precise one, then execute it in the same turn.
 
 ## Hard rules
 
 - Show a diff and wait for explicit approval before changing root instruction files, `backbone.yml`, rules, skills, or workflows.
 - Do not deploy, rotate secrets, run migrations, delete data, or rewrite remote history without explicit approval.
+- Prefer `trash` over `rm` for deletions; permanent deletes need explicit approval of the exact paths (see `.claude/rules/safe-delete.md`).
 - Keep `.autoresearch/`, `results.tsv`, and `.vibekit/reports/` local unless the user asks to commit them.
 
 ## English Learning & Grammar Correction

package/.vibekit/init/FIRST_PROMPT.md

@@ -36,7 +36,7 @@ Use the autoresearch-coding skill.
 Goal: improve this Minimal Vibe Coding Kit for existing projects in any language.
 Metric command: node .vibekit/scripts/validate-kit.mjs .
 Direction: higher.
-Editable paths: README.md .vibekit/docs .vibekit/scripts skills .vibekit/commands .claude .cursor .agents .codex-plugin .github backbone.yml AGENTS.md .vibekit/init/CLAUDE-template.md .vibekit/init/FIRST_TIME_INIT.md package.json install.sh install.ps1.
+Editable paths: README.md .vibekit/docs .vibekit/scripts skills .vibekit/commands .claude .cursor .agents .grok .codex-plugin .github backbone.yml AGENTS.md .vibekit/init/CLAUDE-template.md .vibekit/init/FIRST_TIME_INIT.md package.json install.sh install.ps1.
 Protected paths: .git .env* node_modules vendor secrets lockfiles.
 Budget: 3.
 ```

package/.vibekit/init/FIRST_TIME_INIT.md

@@ -1,6 +1,6 @@
 # .vibekit/init/FIRST_TIME_INIT.md - bootstrap runbook
 
-Audience: Claude, Cursor, Codex, or any AI coding agent in a repo that just received Minimal Vibe Coding Kit.
+Audience: Claude, Cursor, Codex, Grok, or any AI coding agent in a repo that just received Minimal Vibe Coding Kit.
 
 Keep this flow short. Detailed schema notes live in `.vibekit/docs/BACKBONE_REFERENCE.md`.
 
@@ -11,7 +11,7 @@ Before writing anything, print this checklist and mark each item as found, missi
 1. `backbone.yml` exists.
 2. `AGENTS.md` exists or will be created/merged.
 3. `CLAUDE.md` exists or will be created from `.vibekit/init/CLAUDE-template.md` when Claude support is installed.
-4. At least one harness folder exists: `.claude/`, `.cursor/`, `.agents/`, or `.codex-plugin/`.
+4. At least one harness folder exists: `.claude/`, `.cursor/`, `.agents/`, `.grok/`, or `.codex-plugin/`.
 5. Git status is known. If user changes exist, do not overwrite them.
 6. A validation command is inferred or set to `node .vibekit/scripts/validate-kit.mjs .` for the kit itself.
 7. Protected paths include `.env*`, secrets, lockfiles, generated files, build output, migrations, and infra unless the user says otherwise.
@@ -69,6 +69,38 @@ Then:
 - Record domain terms and acronyms in `.vibekit/docs/CONTEXT.md` (scaffold from `.vibekit/docs/templates/CONTEXT_TEMPLATE.md`) and link it in `project.context`, so future agents stay concise.
 - Keep the PRD short: it captures intent and focus, not a full specification.
 
+## Setup preferences
+
+Ask these two questions right after the PRD interview. Persist both answers as `backbone.yml` `conventions.custom_rules` entries in the same proposed diff.
+
+### 1. Safe delete (trash instead of rm)
+
+1. Detect first: run `command -v trash` (read-only).
+2. If `trash` is available, ask: "Use `trash` instead of `rm` for deletions so they are recoverable? (recommended: yes)"
+3. If `trash` is missing, recommend installing it, then ask the same question:
+   - macOS 14 or newer: already built in at `/usr/bin/trash`, nothing to install.
+   - Older macOS: `brew install trash` (https://formulae.brew.sh/formula/trash)
+   - Linux: `sudo apt install trash-cli` or `pip install trash-cli` (https://github.com/andreafrancia/trash-cli)
+   - Any OS with Node.js: `npm install --global trash-cli` (https://github.com/sindresorhus/trash-cli)
+4. Record the answer as a custom rule:
+   - yes: `Deletion safety: use trash instead of rm; permanent deletes require explicit approval of the exact paths.`
+   - no: `Deletion safety: user declined trash; list the exact paths and wait for approval before any rm.`
+
+### 2. Default coding level
+
+Show this table exactly and ask: "Which explanation level should I use by default? Reply 0-5. You can change it anytime with `/coding-level N`."
+
+| N | Level | Short description |
+| --- | --- | --- |
+| 0 | ELI5 | New to code. Everything in plain words, no jargon. |
+| 1 | Junior | Knows the basics. Explains the why behind every step. |
+| 2 | Mid-level | Comfortable coder. Patterns, architecture, trade-offs. |
+| 3 | Senior | Concise. Trade-offs, edge cases, operational concerns. |
+| 4 | Tech Lead | Strategic. Risk, ROI, team and organizational impact. |
+| 5 | God | Expert peer. Minimal explanation, maximum signal. |
+
+Record the answer as a custom rule: `Default coding level: N (<Level>) - apply this explanation depth every session; change with /coding-level N.`
+
 ## FIRST_TIME_INIT_RULES
 
 Create project rules from evidence in the existing repo, not from a fixed language or framework template.

package/.vibekit/scripts/doctor.mjs

@@ -153,6 +153,13 @@ function statusLine(ok, label, detail) {
   return `${ok ? 'PASS' : 'WARN'} ${label}${detail ? ` - ${detail}` : ''}`;
 }
 
+const TRASH_INSTALL_HINT = 'macOS 14+ has it built in; older macOS: brew install trash; Linux: sudo apt install trash-cli; any OS with Node: npm i -g trash-cli';
+
+function trashAvailable() {
+  const result = spawnSync('trash', ['--help'], { encoding: 'utf8' });
+  return !(result.error && result.error.code === 'ENOENT');
+}
+
 const backbone = parseBackbone();
 const packageJson = readJson('package.json');
 const isKitTemplate = packageJson?.name === 'minimal-vibe-coding-kit';
@@ -163,7 +170,8 @@ const surfaces = {
   agents: exists('AGENTS.md'),
   claude: exists('CLAUDE.md') || exists('.vibekit/init/CLAUDE-template.md') || exists('.claude'),
   cursor: exists('.cursor/rules'),
-  codex: exists('.agents') || exists('.codex-plugin/plugin.json')
+  codex: exists('.agents') || exists('.codex-plugin/plugin.json'),
+  grok: exists('.grok')
 };
 const managedBlocks = {
   AGENTS: countManaged('AGENTS.md'),
@@ -183,7 +191,8 @@ const nativeSkillSurfaces = {
   shared: '.vibekit/skills',
   claude: '.claude/skills',
   codex: '.agents/skills',
-  cursor: '.cursor/skills'
+  cursor: '.cursor/skills',
+  grok: '.grok/skills'
 };
 const missingNativeSkills = [];
 for (const [surface, dir] of Object.entries(nativeSkillSurfaces)) {
@@ -194,8 +203,11 @@ for (const [surface, dir] of Object.entries(nativeSkillSurfaces)) {
 const commands = commandMap();
 const validation = runNodeScript('.vibekit/scripts/validate-kit.mjs');
 const probe = runProbe();
+const hasTrash = trashAvailable();
 const risks = [];
 
+if (!hasTrash) risks.push(`trash command not found; agents fall back to asking before rm. Install it for recoverable deletes (${TRASH_INSTALL_HINT}).`);
+
 if (!backbone) risks.push('backbone.yml is missing.');
 else if (backbone.templateStatus !== 'initialized' && !isKitTemplate) risks.push(`backbone.yml is ${backbone.templateStatus || 'unknown'}; run init before relying on it.`);
 if (!Object.values(surfaces).some(Boolean)) risks.push('No AI agent surfaces were detected.');
@@ -238,12 +250,18 @@ const report = {
     command: probe.command,
     status: probe.status
   },
+  safeDelete: {
+    trashAvailable: hasTrash,
+    installHint: hasTrash ? null : TRASH_INSTALL_HINT
+  },
   aiRulesLoaded: {
     sharedSkills: listFiles('.vibekit/skills').filter((f) => f.endsWith('SKILL.md')).length,
     claudeSkills: listFiles('.claude/skills').filter((f) => f.endsWith('SKILL.md')).length,
     codexSkills: listFiles('.agents/skills').filter((f) => f.endsWith('SKILL.md')).length,
     cursorSkills: listFiles('.cursor/skills').filter((f) => f.endsWith('SKILL.md')).length,
-    cursorRules: listFiles('.cursor/rules').filter((f) => f.endsWith('.mdc')).length
+    cursorRules: listFiles('.cursor/rules').filter((f) => f.endsWith('.mdc')).length,
+    grokSkills: listFiles('.grok/skills').filter((f) => f.endsWith('SKILL.md')).length,
+    grokRules: listFiles('.grok/rules').filter((f) => f.endsWith('.md')).length
   },
   recommendedFirstPrompt: 'Read AGENTS.md and backbone.yml, run mvck doctor ., then propose a small safe plan before editing.',
   knownRisks: risks
@@ -256,7 +274,8 @@ function renderMarkdown(data) {
     statusLine(Boolean(commands.validate), 'validation command detected', commands.validate || 'missing'),
     statusLine(protectedSane, 'protected paths are sane', `${data.protectedPaths.length} entries`),
     statusLine(validation.found && validation.status === 0, 'validation runs', validation.found ? `exit ${validation.status}` : 'missing'),
-    statusLine(probe.found && probe.status === 0, 'AgentShield probe runs', probe.found ? `exit ${probe.status}` : 'missing')
+    statusLine(probe.found && probe.status === 0, 'AgentShield probe runs', probe.found ? `exit ${probe.status}` : 'missing'),
+    statusLine(hasTrash, 'trash command available for safe deletes', hasTrash ? 'deletions are recoverable' : TRASH_INSTALL_HINT)
   ];
 
   return `# Vibe Report
@@ -293,15 +312,18 @@ ${(data.protectedPaths.length ? data.protectedPaths : ['not declared']).map((ite
 - Claude surface: ${data.agentSurfaces.claude ? 'yes' : 'no'}
 - Cursor surface: ${data.agentSurfaces.cursor ? 'yes' : 'no'}
 - Codex surface: ${data.agentSurfaces.codex ? 'yes' : 'no'}
+- Grok surface: ${data.agentSurfaces.grok ? 'yes' : 'no'}
 - Shared skills: ${data.aiRulesLoaded.sharedSkills}
 - Claude skills: ${data.aiRulesLoaded.claudeSkills}
 - Codex skills: ${data.aiRulesLoaded.codexSkills}
 - Cursor skills: ${data.aiRulesLoaded.cursorSkills}
 - Cursor rules: ${data.aiRulesLoaded.cursorRules}
+- Grok skills: ${data.aiRulesLoaded.grokSkills}
+- Grok rules: ${data.aiRulesLoaded.grokRules}
 
 ## Native reasoning skills
 
-${data.nativeReasoningSkills.names.map((skill) => `- ${skill}: ${data.nativeReasoningSkills.missing.some((item) => item.endsWith(`:${skill}`)) ? 'missing on one or more surfaces' : 'available on shared, Claude, Codex, and Cursor'}`).join('\n')}
+${data.nativeReasoningSkills.names.map((skill) => `- ${skill}: ${data.nativeReasoningSkills.missing.some((item) => item.endsWith(`:${skill}`)) ? 'missing on one or more surfaces' : 'available on shared, Claude, Codex, Cursor, and Grok'}`).join('\n')}
 
 ## Recommended first prompt
 

package/.vibekit/scripts/init-backbone.mjs

@@ -1,8 +1,14 @@
 #!/usr/bin/env node
 import fs from 'node:fs';
 import path from 'node:path';
+import { spawnSync } from 'node:child_process';
 
 const args = process.argv.slice(2);
+
+function trashAvailable() {
+  const result = spawnSync('trash', ['--help'], { encoding: 'utf8' });
+  return !(result.error && result.error.code === 'ENOENT');
+}
 const shouldWrite = args.includes('--write');
 const yes = args.includes('--yes');
 
@@ -619,6 +625,9 @@ if (shouldWrite) {
   console.log(inferred.context === 'none'
     ? '- Offer to create .vibekit/docs/CONTEXT.md (glossary) from .vibekit/docs/templates/CONTEXT_TEMPLATE.md.'
     : `- CONTEXT glossary found at ${inferred.context}; keep it current.`);
+  console.log('\nSetup preferences (record answers in conventions.custom_rules; see FIRST_TIME_INIT.md "Setup preferences"):');
+  console.log(`- Safe delete: trash command ${trashAvailable() ? 'available' : 'NOT found (recommend: macOS 14+ built-in; older macOS `brew install trash`; Linux `sudo apt install trash-cli`; any OS `npm i -g trash-cli`)'} - ask: use trash instead of rm? (recommended: yes)`);
+  console.log('- Default coding level: ask 0-5 (0 ELI5, 1 Junior, 2 Mid-level, 3 Senior, 4 Tech Lead, 5 God); changeable later with /coding-level N.');
   console.log('\nConvention review questions:');
   for (const question of inferred.conventions.reviewQuestions) console.log(`- ${question}`);
   console.log('\nProposed backbone.yml:\n');

package/.vibekit/scripts/mvck.mjs

@@ -13,8 +13,8 @@ function usage() {
   console.log(`Minimal Vibe Coding Kit
 
 Usage:
-  mvck install [target] [--profile all|claude,cursor,codex] [--force] [--dry-run] [--json]
-  mvck update [target] [--profile all|claude,cursor,codex] [--dry-run] [--json] [--no-backup]
+  mvck install [target] [--profile all|claude,cursor,codex,grok] [--force] [--dry-run] [--json]
+  mvck update [target] [--profile all|claude,cursor,codex,grok] [--dry-run] [--json] [--no-backup]
   mvck init [target] [--propose|--write --yes] [--preset nextjs|wordpress|python|laravel|docker]
   mvck validate [target]
   mvck doctor [target] [--write-report] [--json]
@@ -158,23 +158,24 @@ const KIT_SCRIPTS = [
   '.vibekit/scripts/mvck.mjs', '.vibekit/scripts/init-backbone.mjs', '.vibekit/scripts/daily-enhance.mjs', '.vibekit/scripts/validate-kit.mjs',
   '.vibekit/scripts/doctor.mjs', '.vibekit/scripts/agentshield-probe.mjs', '.vibekit/scripts/vibekit-finalize.mjs'
 ];
-const VALID_PROFILES = new Set(['claude', 'cursor', 'codex']);
+const VALID_PROFILES = new Set(['claude', 'cursor', 'codex', 'grok']);
 
 function parseProfiles(profileRaw) {
-  const profiles = new Set(profileRaw === 'all' ? ['claude', 'cursor', 'codex'] : profileRaw.split(',').map((x) => x.trim()).filter(Boolean));
+  const profiles = new Set(profileRaw === 'all' ? ['claude', 'cursor', 'codex', 'grok'] : profileRaw.split(',').map((x) => x.trim()).filter(Boolean));
   for (const p of profiles) {
-    if (!VALID_PROFILES.has(p)) throw new Error(`Unknown profile: ${p}. Valid values: all, claude, cursor, codex (comma-separated).`);
+    if (!VALID_PROFILES.has(p)) throw new Error(`Unknown profile: ${p}. Valid values: all, claude, cursor, codex, grok (comma-separated).`);
   }
   return profiles;
 }
 const CLAUDE_DIRS = ['.claude/agents', '.claude/commands', '.claude/rules'];
 const CLAUDE_SKILLS = [
   'autoresearch-coding', 'agentshield-security-review', 'daily-workflow-curator', 'vibekit-init', 'visual-design-loop',
-  'clearthought', 'sequential-thinking', 'reviewing-4p-priorities', 'path-sensitive-shell-safety', 'memento', 'coding-level', 'parallel-analysis'
+  'clearthought', 'sequential-thinking', 'reviewing-4p-priorities', 'path-sensitive-shell-safety', 'memento', 'coding-level', 'parallel-analysis', 'prompt-sharpener'
 ];
 const CURSOR_DIRS = ['.cursor/rules', '.cursor/commands'];
-const CURSOR_SKILLS = ['clearthought', 'sequential-thinking', 'reviewing-4p-priorities', 'path-sensitive-shell-safety', 'memento', 'coding-level', 'parallel-analysis'];
+const CURSOR_SKILLS = ['clearthought', 'sequential-thinking', 'reviewing-4p-priorities', 'path-sensitive-shell-safety', 'memento', 'coding-level', 'parallel-analysis', 'prompt-sharpener'];
 const CODEX_DIRS = ['.agents', '.codex', '.codex-plugin'];
+const GROK_DIRS = ['.grok'];
 const GITIGNORE_BLOCK = `# BEGIN: minimal-vibe-coding-kit\n.autoresearch/\nresults.tsv\n.vibekit/INIT_DONE\n.vibekit/FINALIZE_DONE\n.vibekit/reports/\n.vibekit/update-backup/\n_vibekit-cleanup/\nCLAUDE.local.md\n# END: minimal-vibe-coding-kit`;
 
 function kitVersion() {
@@ -258,12 +259,18 @@ function install() {
       actions.push(copyDirSafe(`.cursor/skills/${skill}`, `.cursor/skills/${skill}`, target, opts));
     }
     actions.push(copyFileSafe('.cursor/settings.json', '.cursor/settings.json', target, opts));
+    actions.push(copyFileSafe('.cursor/cli.json', '.cursor/cli.json', target, opts));
   }
   if (profiles.has('codex')) {
     for (const dir of CODEX_DIRS) {
       actions.push(copyDirSafe(dir, dir, target, opts));
     }
   }
+  if (profiles.has('grok')) {
+    for (const dir of GROK_DIRS) {
+      actions.push(copyDirSafe(dir, dir, target, opts));
+    }
+  }
 
   applyManagedBlocks(target, profiles, actions, { dryRun });
   writeKitVersion(target, dryRun);
@@ -398,13 +405,21 @@ function update() {
   if (profiles.has('codex')) {
     for (const dir of CODEX_DIRS) actions.push(...updateDirSafe(dir, target, opts));
   }
+  if (profiles.has('grok')) {
+    // .grok/config.toml holds user-editable permission rules; seed it below instead of overwriting.
+    for (const dir of GROK_DIRS) actions.push(...updateDirSafe(dir, target, dir === '.grok' ? { ...opts, exclude: ['config.toml'] } : opts));
+  }
 
   // User-owned files: seed only when missing, never overwrite. Finalized
   // projects removed the one-time files on purpose, so skip re-seeding them.
   const seedFiles = finalized ? ['backbone.yml'] : KIT_SEED_FILES;
   for (const file of seedFiles) actions.push(copyFileSafe(file, file, target, { force: false, dryRun }));
   if (profiles.has('claude')) actions.push(copyFileSafe('.claude/settings.json', '.claude/settings.json', target, { force: false, dryRun }));
-  if (profiles.has('cursor')) actions.push(copyFileSafe('.cursor/settings.json', '.cursor/settings.json', target, { force: false, dryRun }));
+  if (profiles.has('cursor')) {
+    actions.push(copyFileSafe('.cursor/settings.json', '.cursor/settings.json', target, { force: false, dryRun }));
+    actions.push(copyFileSafe('.cursor/cli.json', '.cursor/cli.json', target, { force: false, dryRun }));
+  }
+  if (profiles.has('grok')) actions.push(copyFileSafe('.grok/config.toml', '.grok/config.toml', target, { force: false, dryRun }));
 
   applyManagedBlocks(target, profiles, actions, { dryRun });
   writeKitVersion(target, dryRun);
@@ -412,7 +427,7 @@ function update() {
   const summary = { add: 0, update: 0, unchanged: 0, skip: 0, 'managed-block': 0 };
   for (const a of actions) summary[a.action] = (summary[a.action] || 0) + 1;
   const backupInfo = backup && backup.count > 0 ? path.relative(target, backup.dir) : null;
-  const preserved = ['backbone.yml', 'CLAUDE.md', 'AGENTS.md content outside the managed block', '.claude/settings.json', '.cursor/settings.json'];
+  const preserved = ['backbone.yml', 'CLAUDE.md', 'AGENTS.md content outside the managed block', '.claude/settings.json', '.cursor/settings.json', '.cursor/cli.json', '.grok/config.toml'];
 
   const legacyPaths = detectLegacyLayout(target);
   if (json) {

package/.vibekit/scripts/test-install.mjs

@@ -120,6 +120,14 @@ try {
   const updParsed = JSON.parse(updPlan.stdout);
   assert(updParsed.status === 'dry-run' && typeof updParsed.toVersion === 'string', 'update --dry-run --json returns machine-readable plan');
 
+  // Single-profile installs must pass validation on their own.
+  for (const profile of ['claude', 'cursor', 'codex', 'grok']) {
+    const solo = tempDir(`profile-${profile}`);
+    run(['.vibekit/scripts/mvck.mjs', 'install', solo, '--profile', profile]);
+    run(['.vibekit/scripts/validate-kit.mjs', solo]);
+    assert(true, `${profile}-only install passes validation`);
+  }
+
   console.log('\nInstall behavior tests passed.');
 } finally {
   if (!keep) {