raditia/CraftKit
One repo of AI coding skills that auto-syncs across Claude Code, Cursor, Gemini CLI, and Codex CLI. Pull once — every AI tool gets the same workflows, rules, and commands.
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
Copy this rig
# review before running: this installs third-party code $ npx degit raditia/CraftKit/agents ./rig-craftkit/agents $ npx degit raditia/CraftKit/commands ./rig-craftkit/commands $ npx degit raditia/CraftKit/skills ./rig-craftkit/skills $ npx degit raditia/CraftKit/hooks ./rig-craftkit/hooks
MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ npx degit raditia/CraftKit/skills/adr .claude/skills/adr $ npx degit raditia/CraftKit/skills/android-a11y .claude/skills/android-a11y $ npx degit raditia/CraftKit/skills/android-context .claude/skills/android-context $ npx degit raditia/CraftKit/skills/android-patterns .claude/skills/android-patterns $ npx degit raditia/CraftKit/skills/android-performance .claude/skills/android-performance $ npx degit raditia/CraftKit/skills/android-review .claude/skills/android-review $ npx degit raditia/CraftKit/skills/android-scaffold .claude/skills/android-scaffold $ npx degit raditia/CraftKit/skills/android-test .claude/skills/android-test $ npx degit raditia/CraftKit/skills/code-quality .claude/skills/code-quality $ npx degit raditia/CraftKit/skills/context-source .claude/skills/context-source $ npx degit raditia/CraftKit/skills/debug .claude/skills/debug $ npx degit raditia/CraftKit/skills/docs .claude/skills/docs $ npx degit raditia/CraftKit/skills/eval .claude/skills/eval $ npx degit raditia/CraftKit/skills/fe-a11y .claude/skills/fe-a11y $ npx degit raditia/CraftKit/skills/fe-context .claude/skills/fe-context $ npx degit raditia/CraftKit/skills/fe-design .claude/skills/fe-design $ npx degit raditia/CraftKit/skills/fe-patterns .claude/skills/fe-patterns $ npx degit raditia/CraftKit/skills/fe-performance .claude/skills/fe-performance $ npx degit raditia/CraftKit/skills/fe-review .claude/skills/fe-review $ npx degit raditia/CraftKit/skills/fe-scaffold .claude/skills/fe-scaffold $ npx degit raditia/CraftKit/skills/fe-test .claude/skills/fe-test $ npx degit raditia/CraftKit/skills/grill .claude/skills/grill $ npx degit raditia/CraftKit/skills/handoff .claude/skills/handoff $ npx degit raditia/CraftKit/skills/ideate .claude/skills/ideate $ npx degit raditia/CraftKit/skills/interview .claude/skills/interview $ npx degit raditia/CraftKit/skills/ios-a11y .claude/skills/ios-a11y $ npx degit raditia/CraftKit/skills/ios-context .claude/skills/ios-context $ npx degit raditia/CraftKit/skills/ios-patterns .claude/skills/ios-patterns $ npx degit raditia/CraftKit/skills/ios-performance .claude/skills/ios-performance $ npx degit raditia/CraftKit/skills/ios-review .claude/skills/ios-review $ npx degit raditia/CraftKit/skills/ios-scaffold .claude/skills/ios-scaffold $ npx degit raditia/CraftKit/skills/ios-test .claude/skills/ios-test $ npx degit raditia/CraftKit/skills/plan .claude/skills/plan $ npx degit raditia/CraftKit/skills/ponytail-audit .claude/skills/ponytail-audit $ npx degit raditia/CraftKit/skills/ponytail-debt .claude/skills/ponytail-debt $ npx degit raditia/CraftKit/skills/ponytail-review .claude/skills/ponytail-review $ npx degit raditia/CraftKit/skills/research .claude/skills/research $ npx degit raditia/CraftKit/skills/spec .claude/skills/spec $ npx degit raditia/CraftKit/skills/test-cases .claude/skills/test-cases $ npx degit raditia/CraftKit/skills/think .claude/skills/think
$ curl -fsSL --create-dirs -o .claude/agents/adversarial.md https://raw.githubusercontent.com/raditia/CraftKit/main/agents/adversarial.md $ curl -fsSL --create-dirs -o .claude/agents/android-a11y.md https://raw.githubusercontent.com/raditia/CraftKit/main/agents/android-a11y.md $ curl -fsSL --create-dirs -o .claude/agents/android-performance.md https://raw.githubusercontent.com/raditia/CraftKit/main/agents/android-performance.md $ curl -fsSL --create-dirs -o .claude/agents/android-review.md https://raw.githubusercontent.com/raditia/CraftKit/main/agents/android-review.md $ curl -fsSL --create-dirs -o .claude/agents/bulk-read.md https://raw.githubusercontent.com/raditia/CraftKit/main/agents/bulk-read.md $ curl -fsSL --create-dirs -o .claude/agents/code-quality.md https://raw.githubusercontent.com/raditia/CraftKit/main/agents/code-quality.md $ curl -fsSL --create-dirs -o .claude/agents/eval-judge.md https://raw.githubusercontent.com/raditia/CraftKit/main/agents/eval-judge.md $ curl -fsSL --create-dirs -o .claude/agents/fe-a11y.md https://raw.githubusercontent.com/raditia/CraftKit/main/agents/fe-a11y.md $ curl -fsSL --create-dirs -o .claude/agents/fe-patterns.md https://raw.githubusercontent.com/raditia/CraftKit/main/agents/fe-patterns.md $ curl -fsSL --create-dirs -o .claude/agents/fe-performance.md https://raw.githubusercontent.com/raditia/CraftKit/main/agents/fe-performance.md $ curl -fsSL --create-dirs -o .claude/agents/fe-review.md https://raw.githubusercontent.com/raditia/CraftKit/main/agents/fe-review.md $ curl -fsSL --create-dirs -o .claude/agents/ios-a11y.md https://raw.githubusercontent.com/raditia/CraftKit/main/agents/ios-a11y.md $ curl -fsSL --create-dirs -o .claude/agents/ios-performance.md https://raw.githubusercontent.com/raditia/CraftKit/main/agents/ios-performance.md $ curl -fsSL --create-dirs -o .claude/agents/ios-review.md https://raw.githubusercontent.com/raditia/CraftKit/main/agents/ios-review.md $ curl -fsSL --create-dirs -o .claude/agents/plan-roaster.md https://raw.githubusercontent.com/raditia/CraftKit/main/agents/plan-roaster.md $ curl -fsSL --create-dirs -o .claude/agents/ponytail-review.md https://raw.githubusercontent.com/raditia/CraftKit/main/agents/ponytail-review.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 7,204 rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(rm -rf *)",
"Read(**/*.pem)",
"Bash(rm -rf /)",
"Bash(git push --force:*)",
"Bash(sudo *)",
"Read(.env)",
"Bash(rm -rf /*)",
"Read(./.env.*)",
"Read(~/.aws/**)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(sudo:*)",
"Bash(git reset --hard*)",
"Bash(git reset --hard:*)",
"Read(.env.*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(rm:*)",
"Bash(git rebase *)",
"Bash(wget *)",
"Bash(npm publish:*)",
"Bash(git commit *)",
"Bash(gh pr merge *)"
]
}
} Skills (40)
adrandroid-a11yandroid-contextandroid-patternsandroid-performanceandroid-reviewandroid-scaffoldandroid-testcode-qualitycontext-sourcedebugdocsevalfe-a11yfe-contextfe-designfe-patternsfe-performancefe-reviewfe-scaffoldfe-testgrillhandoffideateinterviewios-a11yios-contextios-patternsios-performanceios-reviewios-scaffoldios-testplanponytail-auditponytail-debtponytail-reviewresearchspectest-casesthink
Subagents (16)
| adversarial model: sonnet | Devil's advocate reviewer. Spawned by parallel workflows when 3+ EVPMR layers change or a large module is built, to argue the strongest case against merging. Never edits files. |
| android-a11y model: sonnet | Cold accessibility reviewer for Android (Views + Data Binding and Jetpack Compose). Spawned by parallel workflows when an Activity/Fragment/Widget/Composable or layout XML changes, receiving diff or f |
| android-performance model: sonnet | Cold performance reviewer for Android (Views + RecyclerView and Jetpack Compose) within MVP. Spawned by parallel workflows when a Presenter, ViewModel, adapter, or Composable changes, receiving diff o |
| android-review model: sonnet | Cold Android MVP + Core-framework pattern checker. Spawned by parallel workflows when *.kt/*.java changes, receiving diff or file content inline. Checks layer violations, Dagger DI, NavigatorService n |
| bulk-read model: sonnet | Cold single-shot file reader. Spawned when a whole-file Read is gated for size, or whenever a question about a large file matters more than its text. Reads the paths it is given, answers one question, |
| code-quality model: sonnet | Cold 5-axis code reviewer (correctness, readability, architecture, security, performance). Spawned by parallel workflows, receiving diff and context inline. Never edits files. |
| eval-judge model: sonnet | Cold LLM-as-judge that scores a finished agentic run into a weighted correctness percentage. Spawned by /eval (and offered at the tail of parallel-build / parallel-ship), receiving the diff, the accep |
| fe-a11y model: sonnet | Cold accessibility reviewer for React Native and Next.js. Spawned by parallel workflows when View*.tsx changes, receiving diff or file content inline. Never edits files. |
| fe-patterns model: sonnet | Cold composition patterns reviewer. Spawned by parallel-build, receiving newly built file content inline. Checks hooks discipline, state location, and component structure. Never edits files. |
| fe-performance model: sonnet | Cold performance reviewer for React Native and Next.js. Spawned by parallel workflows when View*.tsx or Presenter*.ts changes, receiving diff or file content inline. Never edits files. |
| fe-review model: sonnet | Cold EVPMR pattern checker. Spawned by parallel workflows, receiving diff or file content inline. Checks layer violations, TypeScript, styling, React correctness, and tracking placement. Never edits f |
| ios-a11y model: sonnet | Cold accessibility reviewer for UIKit-based iOS screens within MVVM-C. Spawned by parallel workflows when a ViewController or View changes, receiving diff or file content inline. Never edits files. |
| ios-performance model: sonnet | Cold performance reviewer for UIKit-based iOS screens within MVVM-C. Spawned by parallel workflows when a ViewModel, ViewController, Fetcher, or cell changes, receiving diff or file content inline. Ne |
| ios-review model: sonnet | Cold iOS MVVM-C pattern checker. Spawned by parallel workflows when *.swift/*.m changes, receiving diff or file content inline. Checks layer violations, Dependency-struct DI, Coordinator-only navigati |
| plan-roaster model: sonnet | Use proactively to stress-test any plan or design doc before implementation. Reads the plan, hunts for the weakest assumption, returns a short verdict. Use when a plan looks finished and you want an u |
| ponytail-review model: sonnet | Cold over-engineering reviewer. Spawned by parallel workflows when non-test source changes, to find what to delete: reinvented stdlib, speculative abstractions, dead flexibility. Never edits files. Co |
Slash commands (11)
/build/cross-review/define/fix/parallel-build/parallel-review/parallel-ship/pr-message/review/ship/team-build
Similar rigs
latuconsinafr/claude-code-config
The claude code config, tailored to myself, battle-tested (might inspires you)
Skill Collector 6.1k tok ·
phnx-labs/.agents
Curated upstream config for agents-cli users
Skill Collector 9.0k tok ·
iamvonpasion/hashb
Dev workflow toolkit for Claude Code — 14 skills that turn AI into a structured development partner. Spec → Design → Eng → TDD → Review → QA → Ship.
Skill Collector 2.5k tok ·
rness-dev/rness
One source of truth for every AI coding agent. The standards, decisions, specs and plans of your organization, written into the AGENTS.md of every repository. Claude Code, Codex, Cursor and GitHub Copilot already read it.
Automator 5.6k tok ·