A collection of custom skills for Claude Code — specialized agents that work standalone or chain together as orchestrated pipelines to validate, plan, architect, and launch tech products.
Built and maintained by @RashiD2801. Calibrated for India-first builders, but useful anywhere.
Skills are prompt files that extend Claude Code with custom slash commands. Drop a skill into your .claude/skills/ folder and it becomes available as /skill-name in any Claude Code session.
Each skill in this library is a self-contained agent with defined tools, steps, and output formats. Some skills are standalone; others are designed to plug into a larger orchestration pipeline.
Clone this repo and copy the skills you want into your Claude Code skills directory:
git clone https://github.com/RashiD2801/AI-Skills.git
# Copy a single skill
cp -r AI-Skills/concept-explorer ~/.claude/skills/
# Or copy everything
cp -r AI-Skills/* ~/.claude/skills/Then restart Claude Code (or reload the window). Skills appear as /skill-name in your session.
Windows users: Your Claude Code skills directory is typically
C:\Users\<you>\.claude\skills\
A full startup validation and build pipeline — run individually or let startup-analyzer orchestrate everything.
| Skill | Command | What it does |
|---|---|---|
| Startup Analyzer | /startup-analyzer "your idea" |
Master orchestrator — runs all 8 specialist agents in order with 2 user checkpoints, produces a full tabbed HTML report |
| Market Research | /startup-market-research |
TAM/SAM/SOM breakdown (India primary), problem statement, ICP profiles, gap analysis, 2–3 pivot options |
| Competitor Scan | /startup-competitor-scan |
Live web search for real global + India-specific competitors, feature gaps, pricing, user complaints, whitespace |
| Financial Feasibility | /startup-financial-feasibility |
MVP cost (India rates), team & burn rate, revenue models, Year 1–3 projections, break-even (INR primary) |
| Tech Feasibility | /startup-tech-feasibility |
Complexity score, build-vs-buy analysis, technical risks, team requirements, MVP timeline, GREEN/AMBER/RED verdict |
| Tech Stack | /startup-tech-stack |
Opinionated frontend/backend/DB/AI-ML/infra recommendations with justifications and cost estimates |
| Go-to-Market | /startup-go-to-market |
India-first launch plan, MVP scope, pricing strategy, distribution channels, phased expansion roadmap |
| Architecture Diagram | /startup-architecture-diagram |
Self-contained HTML/CSS/SVG architecture diagram showing all components, services, and data flows |
| Report Maker | /startup-report-maker |
Assembles all outputs into a beautiful 8-tab interactive HTML report with a BUILD/PIVOT/PASS verdict card |
| Dev Kickstart | /startup-dev-kickstart |
Generates README, PRD, CLAUDE.md, ARCHITECTURE.md, STACK.md, ROADMAP.md — then creates and pushes a GitHub repo |
| Pitch Deck | /startup-pitch-deck |
Investor or user pitch deck (.pptx) — 15–30 slides with data, charts, financials, and a clean design system |
| Skill | Command | What it does |
|---|---|---|
| Software Developer | /software-developer "what to build" |
Skip business validation — goes straight to tech feasibility → stack → architecture → dev kickstart. For builders who already know what they're building. |
| System Architect | /system-architect |
Interview-driven: describe your system, get a self-contained HTML file with an SVG flow diagram and tech stack recommendation |
| Skill | Command | What it does |
|---|---|---|
| Concept Explorer | /concept-explorer "concept" |
Generates a self-contained HTML study page with explanations, analogies, curated videos/articles, real-world tools, and hands-on app ideas. Saved to D:\AI\Learning\ |
/startup-analyzer "your idea"
│
├── Phase 1 (parallel)
│ ├── Market Research ← TAM, ICP, pivots
│ └── Competitor Scan ← live web search
│
├── Checkpoint 1 — proceed / pivot / stop?
│
├── Phase 2 (parallel)
│ ├── Financial Feasibility ← costs, projections, break-even
│ └── Tech Feasibility ← complexity, risks, verdict
│
├── Checkpoint 2 — proceed / pivot / stop?
│
├── Phase 3 (sequential)
│ ├── Tech Stack ← specific technology choices
│ ├── Go-to-Market ← launch strategy
│ └── Architecture Diagram ← system diagram
│
├── Phase 4 — Report Maker ← tabbed HTML report
├── Phase 5 — Dev Kickstart ← GitHub repo + docs
└── Phase 6 (optional) — Pitch Deck
Each phase agent also works standalone — you can run just /startup-market-research or just /startup-tech-stack if you only need a slice of the analysis.
All generated files land in predictable locations:
| Output | Location |
|---|---|
| Startup reports | D:\AI\Startups\[slug]-[date].html |
| Dev repo docs | D:\AI\Startups\[slug]\ |
| Pitch decks | D:\AI\Startups\[slug]\[slug]-pitch-deck-[date].pptx |
| Learning pages | D:\AI\Learning\learn-[concept].html |
These paths are Windows-specific and match my local setup. If you're on Mac/Linux, edit the
Writesteps in eachSKILL.mdto use your preferred paths.
All generated HTML reports and pages use a consistent visual language:
- Font: DM Sans (Google Fonts)
- Background:
#F9F6F1(off-white) - Primary text:
#1B3A2D(deep forest green) - Accent:
#4A7C6F(muted teal) - Highlight:
#F0A882(peach) - CTA:
#E8724A(strong orange)
Each skill lives in its own folder:
skill-name/
└── SKILL.md ← the skill definition (frontmatter + step-by-step instructions)
SKILL.md uses YAML frontmatter to declare the skill name, description, and allowed tools, followed by step-by-step instructions for the agent.
MIT — use freely, modify to fit your context, and share improvements.