Source of truth: The GitHub repository is the canonical source for Skills Tree. This README is the public source guide and entry point. Current repository state, decisions, roadmap, architecture, quality reports, skills, issues, pull requests, releases, and security policies are maintained in GitHub. No external dashboard or deployment platform is authoritative.
Skills Tree
📆 This Week's Highlights — October 3, 2026
The current corpus contains 375 skill files, including one intentional test fixture classified separately from production skills.
The Public AI Skills Source
Skills Tree is a public, trusted, machine-discoverable source of AI skills.
AI agents should be able to find the right skill here, and humans should be able to discover, understand, use, share, and contribute skills easily.
The repository remains an evidence-backed, dependency-aware knowledge graph and distribution system for making those skills discoverable and reusable.
| 375 Skill Files | 780+ Connections | MIT Licensed |
|---|---|---|
| Versioned & evidence-backed | Dependency-mapped | Community-governed |
Need an AI skill? Start here. Search the machine-readable registry, inspect the canonical skill, then use the Agent Skills package when available.
🔎 Find a Skill · 🤖 Agent Skills · 📚 Browse Canonical Skills · 🧭 AI Discovery Guide · 🧠 llms.txt
▶ Explore Skills → · ▶ Generate Blueprint → · ▶ GitHub Repository →
Response SLA: Issues
<72h· Pull Requests<7 days· Governance Reviews<5 days
AI Repository Governance: Start with AI_CONSTITUTION.md and AGENTS.md before performing repository-level AI work.
📖 README / Source Guide · 📦 PyPI · 🗺️ Systems · 🏗️ Blueprints · 📊 Benchmarks · 🤝 Contribute · 🗺 Roadmap
🌐 Read in your language: 🇬🇧 English · 🇸🇦 العربية · 🇨🇳 中文 · 🇪🇸 Español · 🇩🇪 Deutsch · 🇫🇷 Français · 🇮🇳 हिन्दी · 🇯🇵 日本語 · 🇰🇷 한국어 · 🇧🇷 Português · 🇷🇺 Русский
⚡ Current Usage
The verified repository interfaces are the canonical Markdown skill catalog, generated machine-readable artifacts, the FastAPI service, and the Typer CLI.
For a local checkout:
pip install -e .
skills-tree goals
skills-tree skills
skills-tree recommend --goal "Coding Agent"
skills-tree blueprint --goal "Coding Agent"
skills-tree validate
The historical skills_tree.SkillsTree Python API is not currently implemented in the repository. Search is currently generated from canonical skills/ content into docs/search-index.json for the static web experience. See Python API Status and CLI Reference for the verified boundaries.
→ Full install guide: docs/installation.md · Quick start: docs/quickstart.md
For AI Agents: Use Skills Tree as a Skill Source
When an AI agent needs a reusable capability, treat Skills Tree as a public discovery source:
- Search the machine-readable registry:
docs/api/skills.json. - Match the user's capability or task to a skill by name, description, category, and related skills.
- Open the canonical
skills/<category>/<skill>.mdbefore relying on its instructions. - Check maturity, evidence, dependencies, security boundaries, and freshness.
- Prefer
agent-skills/<name>/SKILL.mdwhen a standards-compatible package exists. - Never treat a registry entry as proof that a skill is safe, current, production-ready, or widely adopted.
- If no matching skill exists, report that gap rather than inventing a Skills Tree entry.
Machine-readable registry: https://raw.githubusercontent.com/SamoTech/skills-tree/main/docs/api/skills.json
Canonical source: https://github.com/SamoTech/skills-tree/tree/main/skills
Agent Skills distribution: https://github.com/SamoTech/skills-tree/tree/main/agent-skills
This is intentionally designed so an AI can discover the project from repository text, resolve a capability deterministically, inspect the canonical source, and then consume a compatible skill package.
The Problem
Every AI agent builder rediscovers the same skills from scratch.
Someone learns RAG the hard way. Someone else figures out memory injection at 2am. A third person spends a week benchmarking ReAct vs LATS — and never shares the results. A fourth discovers the same failure modes you already hit last month.
That collective knowledge is disappearing into Slack threads, private repos, and Twitter bookmarks.
Skills Tree fixes that. → Read the full problem statement
What This Is
Skills Tree is a public source of AI skills built for both AI discovery and human use.
A living, versioned, community-powered collection of AI skills and capability definitions. Skills are documented with practical usage, evidence, failure modes, limitations, and evolution history where supported by repository evidence.
Quality classifications are repository validation states, not proof of production readiness or universal safety. Inspect each skill's evidence, freshness, dependencies, limitations, and security boundaries before use.
→ Real-world use cases · Why Skills Tree vs alternatives
Architecture
┌─────────────────────────────────────────────────────────┐
│ skills-tree │
│ (Python package) │
├─────────────┬───────────────────┬───────────────────────┤
│ CLI │ Python API │ MCP Server │
│ (Typer) │ (FastAPI/Pydantic)│ (tools/mcp/) │
├─────────────┴───────────────────┴───────────────────────┤
│ Skills Data Layer (Markdown + YAML) │
│ skills/ │ systems/ │ blueprints/ │ benchmarks/ │
├─────────────────────────────────────────────────────────┤
│ Validation Engine │ Search Index │ Quality Reports │
│ (tools/) │ (Lunr.js) │ (meta/) │
└─────────────────────────────────────────────────────────┘
→ Full architecture deep-dive: docs/architecture.md
Comparison vs Alternatives
| Feature | Skills Tree | LangChain Hub | Hugging Face Hub | Custom YAML files |
|---|---|---|---|---|
| AI agent skill taxonomy | ✅ 369 skills | ⚠️ Prompt-focused | ❌ Model-focused | ❌ None |
| Versioned skill evolution | ✅ v1→v2→v3 | ❌ | ❌ | ❌ |
| Runnable code examples | ✅ Every skill | ⚠️ Some | ⚠️ Some | ❌ |
| Benchmarks included | ✅ Head-to-head | ❌ | ⚠️ Leaderboards | ❌ |
| MCP server integration | ✅ Built-in | ❌ | ❌ | ❌ |
| Multi-agent blueprints | ✅ 7+ blueprints | ⚠️ Templates | ❌ | ❌ |
| CLI + Python API | ✅ Both | ⚠️ Python only | ✅ Both | ❌ |
| Community-governed | ✅ Open PRs | ⚠️ Curated | ✅ Open | ✅ (yours only) |
| Failure modes documented | ✅ Every skill | ❌ | ❌ | ❌ |
| Free & open source (MIT) | ✅ | ⚠️ Mixed | ✅ | ✅ |
🚀 Start Here — Battle-Tested Skills
If you're new, read these first. Each ships with runnable code, typed I/O, failure modes, and a model-comparison table.
Agent reasoning loops
- ReAct — Thought → Action → Observation, the foundation of tool-using agents
- Chain of Thought — explicit step-by-step reasoning + self-consistency
- Tree of Thought — branched reasoning with scoring + beam search
- Reflection / Reflexion — critique → revise loop on top of any output
- Self-Consistency — sample N chains, majority-vote
- Planning — typed, DAG-validated plans your executor can run
- Task Decomposition — break a goal into atomic, runnable subtasks
Retrieval & memory
- RAG — chunk → embed → retrieve → cite, end-to-end with confidence + threshold
- Vector Store Retrieval — typed top-k cosine search with metadata filtering
- Embedding Generation — batched, content-hash-cached, Matryoshka-truncatable
- Memory Injection — top-K user memories per turn
- Short-Term Memory — token-budgeted rolling window
Calling LLMs in production
- Function / Tool Calling — the primitive that turns an LLM into an agent
- OpenAI API — chat, structured outputs, tools, embeddings, streaming, retry
- Anthropic API — Claude with tool loop, prompt caching, streaming
Code, Web & Security
- Code Generation — spec → AST-validated source with self-repair
- Web Search — Tavily/Serper/Brave with recency + TTL cache
- Input Sanitization — 4-layer defense: structural + boundary + content + isolation
The full battle-tested set is auto-listed in
meta/QUALITY-REPORT.md.
What's Inside
skills-tree/
│
├── skills/ → 374 skill files (classification tracked in the generated quality report)
├── systems/ → Multi-skill workflows (research agent, code reviewer...)
├── blueprints/ → Copy-paste production architectures
├── benchmarks/ → Head-to-head, reproducible skill comparisons
├── labs/ → Experimental & bleeding-edge capabilities
│
├── docs/ → Repository documentation and generated registry artifacts
├── i18n/ → Localized READMEs (10 languages)
├── meta/ → Schema, glossary, frameworks, roadmap, changelog
├── mcp/ → MCP server integration
└── tests/ → pytest test suite
🗂️ The 17 Skill Categories
| # | Category | Skills | What It Covers |
|---|---|---|---|
| 01 | 👁️ Perception | 36 | Text, images, PDFs, code, sensors, databases, screens |
| 02 | 🧠 Reasoning | 46 | Planning, deduction, abduction, causal chains, commonsense |
| 03 | 🗄️ Memory | 19 | Working, episodic, semantic, vector, injection, forgetting |
| 04 | ⚡ Action Execution | 21 | File I/O, HTTP, email, shell, database writes |
| 05 | 💻 Code | 28 | Write, run, debug, review, refactor, test, deploy |
| 06 | 💬 Communication | 15 | Summarize, translate, draft, argue, adapt tone |
| 07 | 🔧 Tool Use | 33 | APIs — GitHub, Slack, Stripe, OpenAI, MCP, A2A |
| 08 | 🎭 Multimodal | 14 | Images, audio, video, VQA, 3D, charts |
| 09 | 🤖 Agentic Patterns | 23 | ReAct, CoT, ToT, MCTS, LATS, RAG, Debate |
| 10 | 🖥️ Computer Use | 20 | Click, type, scroll, OCR, terminal, VM, a11y tree |
| 11 | 🌐 Web | 17 | Search, scrape, crawl, login, fill forms, parse RSS |
| 12 | 📊 Data | 18 | ETL, SQL, embeddings, time series, anomaly detection |
| 13 | 🎨 Creative | 14 | Copywriting, image prompts, SVG, music, scripts |
| 14 | 🔒 Security | 13 | Sandboxing, secret scanning, audit logs, rollback |
| 15 | 🎼 Orchestration | 27 | Multi-agent, state machines, retry, consensus |
| 16 | 🏺 Domain-Specific | 28 | Medical, legal, finance, DevOps, education, science |
| 17 | 🛠️ Infrastructure | 1 | Dependency auditing & supply-chain tooling |
A Skill in 60 Seconds
Every skill file is self-contained and production-ready:
# Memory Injection
Category: memory | Level: intermediate | Stability: stable | Version: v2
## Description
Dynamically inject relevant past memories into an agent's system prompt
before each turn — giving the model user context without filling the window.
## Example
```python
client.messages.create(
system=f"{base_system}\n\n## Memory\n{top_k_memories}",
messages=[{"role": "user", "content": user_message}]
)
```
Every documented battle-tested skill includes: ✅ typed inputs/outputs · ✅ runnable Python code · ✅ frameworks table · ✅ failure modes · ✅ version history
🗺️ Systems — Multi-Skill Workflows
| System | Skills Used | Use Case |
|---|---|---|
| Research Agent | Web search + RAG + Summarize | Deep research automation |
| Coding Agent | Code reading + Write + Debug | End-to-end code generation |
| Code Reviewer | Code reading + Reasoning + Comment gen | Automated PR reviews |
| Data Pipeline Agent | DB reading + ETL + Anomaly detection | Automated data ops |
| Customer Support Bot | Memory injection + Intent + Response gen | Personalized support |
| Computer Use Agent | Screen reading + OCR + Click | Full GUI automation |
🏗️ Blueprints — Production Architectures
| Blueprint | Description |
|---|---|
| RAG Stack | Embed → store → retrieve → generate, fully wired |
| Multi-Agent Workflow | Sequential orchestration with handoffs |
| Multi-Agent Mesh | N specialists + orchestrator, parallel execution |
| Human-in-the-Loop | Approval gates, escalation, audit trails |
| Self-Healing Agent | Error detection, retry logic, rollback |
| Memory-First Agent | Profile + episodic + vector memory combined |
📊 Benchmarks
| Benchmark | Winner | Margin | Link |
|---|---|---|---|
| ReAct vs LATS (HotpotQA) | LATS | +8.3% accuracy | → |
| RAG retrieval strategies | HyDE | +12% recall | → |
| Memory injection methods | Top-K semantic | Best cost/quality | → |
| Function calling comparison | Claude 3.7 | +6% tool accuracy | → |
🤝 How to Contribute
| Type | What It Is | PR Title Format |
|---|---|---|
| New Skill | A capability not yet indexed | feat: add [skill] to [category] |
| Skill Upgrade | Bump v1→v2 with better content | improve: [skill] — v1→v2 |
| Benchmark | Head-to-head with real numbers | benchmark: [skill-a] vs [skill-b] |
| System / Blueprint | Multi-skill workflow or architecture | system: add [name] |
git clone https://github.com/SamoTech/skills-tree.git
cp meta/skill-template.md skills/05-code/my-new-skill.md
# Fill in every section → open a PR
Full guide: CONTRIBUTING.md
🗺️ Roadmap
See the full plan: meta/ROADMAP.md
Near-term (v2.x): Skill dependency graph · Skill Paths · JSON/YAML export · Community ratings
Medium-term (v3.0): LangChain Hub / MCP registry integration · 500+ skills
Long-term: Skills Tree becomes the canonical reference for AI agent capabilities
Product Mission
Skills Tree is a public, trusted, machine-discoverable source of AI skills.
AI agents should be able to find the right skill here, and humans should be able to discover, understand, use, share, and contribute skills easily.
The project grows through genuine utility, reuse, sharing, contribution, and GitHub discovery—not fabricated popularity claims.
⭐ Star this repo · 📦 Install from PyPI · 🔎 Browse Skills · 🤝 Contribute · 💖 Sponsor
The AI Agent Skill OS — built by the community, for the community.
| Dimension | Progress | Score | Weight |
|---|---|---|---|
| 📝 README Quality | ████████░░ |
 |
20% |
| 🔥 Commit Activity | ██████████ |
 |
20% |
| 🌿 Repo Freshness | ██████████ |
|
10% |
| 📚 Documentation | ██████████ |
|
10% |
| ⚙️ CI/CD Setup | ██████████ |
|
10% |
| 🎯 Issue Response | ████████░░ |
 |
10% |
| ⭐ Community Signal | ███░░░░░░░ |
 |
5% |
| 🔀 PR Velocity | ██████████ |
|
10% |
| 🔐 Security | ███░░░░░░░ |
 |
5% |
1 maintainer
Metadata
1 maintainer
Release files for skills-tree 1.74.1
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| skills_tree-1.74.1.tar.gz | 1.6 MB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| skills_tree-1.74.1-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 3.4 MB
Release files / skills_tree-1.74.1.tar.gz
| Download URL | skills_tree-1.74.1.tar.gz |
|---|---|
| Size | 1.6 MB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
ac1b53e48e773d59b6ae2afcfbf279f8f17d1975ed143598c05fde1a60c65206
|
|
BLAKE2b-256 checksum How to use checksums |
020898f78df7322f27c4c84796cb9d1ccd8b7b826b87c87210fa4a8488c4ba43
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Oct 4, 2026.
Transparency logRelease files / skills_tree-1.74.1-py3-none-any.whl
| Download URL | skills_tree-1.74.1-py3-none-any.whl |
|---|---|
| Size | 1.7 MB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
9fc96429fc850a04d820de9edcbd203b8553dc4cff504274783aa6634f68872f
|
|
BLAKE2b-256 checksum How to use checksums |
86fea630189db2bd7d3f43a07278ff35a8d581acb68f8e0541c5fe6f5207a569
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Oct 4, 2026.
Transparency log
