pikakit
v3.9.105
Published
PikaKit — Install 51 skills, 18 workflows, 21 agents into your AI project. One command.
Downloads
5,427
Maintainers
agentskillkitReadme
⚡ PikaKit
Transform your AI Agent into a FAANG-level engineering team
Composable Skills · Coordinated Agents · Intelligent Execution
Install · Skills · Workflows · Agents · Website
🚀 Quick Install
npx pikakitOne command installs 51 skills, 18 workflows, 21 agents, and a complete AI operating system into your project.
What You Get
| Component | Count | Highlights |
|--------------------|:-----------:|-------------------------------------------------------------------------|
| 🧩 Skills | 51 | Architecture, Frontend, Backend, Security, DevOps, AI, Mobile, Testing |
| ⚡ Workflows | 18 | /think · /build · /autopilot · /studio · /validate and more |
| 🤖 Agents | 21 | 16 domain specialists + 5 meta-agents for coordination |
| 📜 Rules | GEMINI.md | AI behavior configuration — the Supreme Law |
| 🛠️ Scripts | 6 | Checklist, verify, preview, session management |
| 💻 CLI | kit | Skill management · agent (optional interactive dashboard) |
🎯 Why PikaKit
⚙️ Auto-Accept
Approve plan once → agent runs to completion with zero interruptions
🧠 Self-Learning
AI remembers mistakes from lessons-learned.yaml and never repeats them
🔧 Custom Skills
Create project-specific skills that teach AI your conventions
| Feature | PikaKit | Generic AI | Other Tools | |--------------------------------|:-----------------:|:-------------:|:-------------:| | ⚡ Auto-Accept Workflow | ✅ | ❌ | ❌ | | 🤖 Multi-Agent Coordination | 21 specialists | 1 generic | 1–3 | | 🧠 Self-Learning Memory | ✅ | ❌ | ❌ | | 💛 100% JavaScript | ✅ | Mixed | Python | | 🛡️ Safety Protocol | TIER -1 | ❌ | ❌ | | 📉 Token Efficiency | Modular loading | Baseline | Varies |
⚡ Quick Start
1. Install
npx pikakitChoose Project (.agent/) or Global (~/.gemini/) scope. Optionally install AutoLearn CLI.
2. Manage Skills
kit list # List all skills
kit info <skill-name> # Show skill details
kit validate # Validate skill structure
kit doctor # System health check3. Run Workflows
/think auth system # Brainstorm 3+ approaches
/plan # Generate detailed plan
/build # Multi-agent implementation
/autopilot # Full autonomous execution
/validate # Run test suite
/studio # UI design with 50+ styles🧩 Skills Catalog (51)
🏗️ Architecture & Planning
| Skill | Description |
|-----------------------------|-------------------------------------------------|
| project-planner | Task planning with dependency graph |
| idea-storm | Socratic questioning & brainstorming |
| lifecycle-orchestrator | End-to-end task lifecycle management |
| system-design | Architecture decisions & ADR |
| context-engineering | Token optimization & agent architecture |
🎨 Frontend & Design
| Skill | Description |
|-----------------------------|-------------------------------------------------|
| react-pro | Modern React patterns, hooks, state management |
| nextjs-pro | App Router, RSC, 60+ optimization rules |
| design-system | Color theory, typography, UX psychology |
| studio | 50+ styles, 97 palettes, 57 font pairings |
| tailwind-kit | Tailwind CSS v4 patterns |
⚙️ Backend & API
| Skill | Description |
|-----------------------------|-------------------------------------------------|
| api-architect | REST, GraphQL, tRPC design |
| nodejs-pro | Hono/Fastify/Express/NestJS patterns |
| python-pro | FastAPI, Django, Flask |
| data-modeler | Schema design, Prisma/Drizzle |
| auth-patterns | OAuth2, JWT, RBAC/ABAC, MFA, Passkeys |
🔐 Security
| Skill | Description |
|-----------------------------|-------------------------------------------------|
| security-scanner | OWASP 2025, supply chain security |
| offensive-sec | Red team tactics, MITRE ATT&CK |
| code-constitution | Constitutional governance (Supreme Law) |
🧪 Testing & Quality
| Skill | Description |
|-----------------------------|-------------------------------------------------|
| test-architect | Unit, integration, E2E strategies |
| e2e-automation | Playwright, visual testing, deep audit |
| code-craft | Clean code, SRP, DRY, KISS |
| code-review | Linting, static analysis, PR review |
| problem-checker | IDE problem detection & auto-fix |
🐛 Debugging
| Skill | Description |
|-----------------------------|-------------------------------------------------|
| debug-pro | 4-phase methodology + defense-in-depth |
| chrome-devtools | Puppeteer CLI, screenshots, Core Web Vitals |
| knowledge-graph | AST parsing, find-usages, impact analysis |
🤖 AI & Agents
| Skill | Description |
|-----------------------------|-------------------------------------------------|
| google-adk-python | Google Agent Development Kit |
| auto-learned | Hierarchical auto-learned patterns |
| auto-learner | Pattern extraction from errors |
| execution-reporter | Agent routing transparency |
| smart-router | Intelligent agent routing |
| mcp-builder | Build MCP servers for AI agents |
| mcp-management | Discover & execute MCP tools |
| ai-artist | Prompt engineering for LLM & image AI |
☁️ DevOps & Infrastructure
| Skill | Description |
|-----------------------------|-------------------------------------------------|
| cicd-pipeline | Safe deployments, rollback strategies |
| gitops | ArgoCD/Flux GitOps workflows |
| git-workflow | Conventional commits, secret detection |
| server-ops | Process management, scaling |
| vercel-deploy | 1-click Vercel deployment |
| observability | OpenTelemetry, unified logs/metrics/traces |
| perf-optimizer | Core Web Vitals, bundle analysis |
📱 Mobile & Specialized
| Skill | Description |
|-----------------------------|-------------------------------------------------|
| mobile-developer | React Native, Flutter, native |
| mobile-design | Mobile-first UI/UX patterns |
| game-development | Game logic & mechanics |
| typescript-expert | Type-level programming, monorepo |
| shell-script | Bash/Linux terminal patterns |
📝 Documentation & Content
| Skill | Description |
|-----------------------------|-------------------------------------------------|
| doc-templates | README, API docs, ADR templates |
| copywriting | AIDA, PAS, conversion formulas |
| seo-optimizer | SEO, E-E-A-T, Core Web Vitals |
| media-processing | FFmpeg, ImageMagick, RMBG |
| agent-browser | AI-optimized browser automation |
📜 Workflows (18)
| Command | Purpose | Agents |
|----------------|----------------------------------------|:---------------:|
| /think | Brainstorm 3+ approaches | planner |
| /plan | Generate detailed PLAN.md | planner, explorer |
| /build | Full-stack implementation | 3–7 |
| /autopilot | Multi-agent autonomous execution | 3+ |
| /cook | Direct implementation from instructions | code-craft |
| /fix | Quick error remediation | debug-pro |
| /validate | Test suite with coverage | qa |
| /diagnose | Root cause debugging | debug |
| /inspect | Defense-in-depth code review | security |
| /studio | UI design (50+ styles, 97 palettes) | frontend |
| /api | API development with OpenAPI docs | backend |
| /mobile | Mobile app development | mobile |
| /game | Game development | gamedev |
| /chronicle | Auto-documentation | docs |
| /launch | Zero-downtime deployment | devops |
| /stage | Dev sandbox & Docker Compose | devops |
| /monitor | Observability setup | devops |
| /optimize | Performance optimization | perf |
🤝 Multi-Agent Coordination (21)
16 domain specialists + 5 meta-agents working together like a FAANG engineering team.
🧠 Meta-Agents (Runtime Control)
| Agent | Role |
|--------------------|-----------------------------------------------------|
| orchestrator | Strategic coordination & execution order |
| assessor | Risk evaluation before risky operations |
| recovery | State management & checkpoint/restore |
| critic | Conflict resolution between agents |
| learner | Continuous improvement from outcomes |
👥 Domain Agents (16)
| Domain | Agent | Key Skills |
|--------------------|----------------------------|----------------------------------------|
| 🎨 Frontend | frontend | react-pro, tailwind-kit, studio |
| ⚙️ Backend | backend | api-architect, nodejs-pro |
| 🗄️ Database | database | data-modeler |
| 🔐 Security | security / pentest | security-scanner, offensive-sec |
| 🧪 Testing | testing / qa | test-architect, e2e-automation |
| ☁️ DevOps | devops | cicd-pipeline, server-ops |
| 📱 Mobile | mobile | mobile-developer, mobile-design |
| 🐛 Debug | debug | debug-pro |
| ⚡ Performance | perf | perf-optimizer |
| 📋 Planning | planner / lead | project-planner, idea-storm |
| 📝 Documentation | docs | doc-templates |
| 🔍 SEO | seo | seo-optimizer |
| 🎮 Game Dev | gamedev | game-development |
| 🔭 Exploration | explorer | knowledge-graph |
| 🏚️ Legacy | legacy | migration patterns |
| 💡 Product | product-lead | requirements, UX |
🛡️ Safety Protocol (TIER -1)
Safety > Recoverability > Correctness > Cleanliness > Convenience
| Rule | Description |
|---------------------------------|------------------------------------------------------|
| 🚫 No Delete | Never delete files without explicit user confirmation |
| ✍️ Safe Modify | Read/Create/Modify allowed; Delete/Overwrite forbidden |
| 🗂️ Git Versioning | All risky changes get git stash or checkpoint commit |
| 🔁 Rollback Guarantee | Previous version always recoverable via Git |
| ⛔ Human Checkpoint | Core logic, auth, config require explicit approval |
| 🛠️ Failure Recovery | 6-level auto-recovery before escalating to user |
🧠 Self-Learning System
AI learns from every mistake and never repeats it.
Error Detected → Root Cause Analysis → Lesson Extracted → lessons-learned.yaml
↓
Future Prevention ← Pattern Matching ← Auto-Learned Patterns ←─┘Teach the AI from your feedback:
User: "This is a bug — you created a new file instead of renaming"
AI: 📚 Learned: [LEARN-003] - Use rename, don't create new simplified filesView learned lessons:
agent # Launch interactive dashboard
agent learn # Teach new patterns
agent recall # Scan for violations
agent stats # View statistics📦 Project Structure
your-project/
├── .agent/
│ ├── GEMINI.md # AI Rules (Supreme Law)
│ │
│ ├── skills/ # 51 Skills
│ │ ├── react-pro/
│ │ ├── debug-pro/
│ │ ├── studio/
│ │ └── ...
│ │
│ ├── workflows/ # 18 Workflows
│ │ ├── think.md
│ │ ├── build.md
│ │ ├── autopilot.md
│ │ └── ...
│ │
│ ├── agents/ # 21 Specialist Agents
│ │ ├── frontend.md
│ │ ├── backend.md
│ │ ├── security.md
│ │ └── ...
│ │
│ └── config/ # Configuration
│
├── kit.cmd / kit # CLI wrapper
└── agent.cmd / agent # AutoLearn wrapper (optional)💻 CLI Reference
kit (Always Installed)
kit list # List all installed skills
kit info <skill-name> # Show skill details
kit validate # Validate skill structure
kit doctor # Check system health
kit cache status # View cache usageagent (Optional — AutoLearn Dashboard)
agent # Launch interactive dashboard
agent learn # Teach new patterns
agent recall # Scan for violations
agent stats # View statistics
agent watch # Real-time monitoringNote: Not installing AutoLearn does not affect workflows, skills, or agents.
🔗 Links
| Resource | URL | |-----------------------------|------------------------------------------------------------------------------------------| | 🌐 Website | pikakit.com | | 📧 Support | [email protected] | | 📦 npm | npmjs.com/package/pikakit | | 🐙 GitHub (Skills) | github.com/pikakit/agent-skills | | 🐙 GitHub (CLI) | github.com/pikakit/add-skill-kit | | 🐛 Issues | github.com/pikakit/agent-skills/issues |
📄 License
Proprietary — See LICENSE for details.
© 2026 PikaKit. All Rights Reserved.
⚡ PikaKit v3.9.105
Composable Skills · Coordinated Agents · Intelligent Execution
pikakit.com · [email protected]
⭐ Star · Install Now · Build Something Great