Skip to content

README.md

Latest commit

 

History

History
40 lines (27 loc) · 3.17 KB

README.md

File metadata and controls

40 lines (27 loc) · 3.17 KB
Raw

Usage Guides

How to load Skillware skills and connect them to language models. Each guide covers one provider adapter in SkillLoader.

Finding skills on disk

SkillLoader.load_skill() accepts an absolute path to a skill directory, or a registry id such as compliance/tos_evaluator. When the id is not already a path on disk, the loader searches in order:

  1. Roots listed in SKILLWARE_SKILL_PATH (OS path separator between multiple roots)
  2. A skills/ directory in the current working directory or its parents
  3. Bundled skills installed with the skillware package (for example under site-packages/skills/)

For pip-installed apps, keep project skills in ./skills/<category>/<name>/ or set SKILLWARE_SKILL_PATH to your skills root.

By default, SkillLoader.load_skill() validates manifest requirements before loading skill.py: unpinned deps must be importable; pinned specifiers (for example web3>=6.0.0) must match the installed version. See Install extras — Loader behavior.

Security: Loading a skill executes its skill.py in your process — there is no sandbox, and the first matching id in the search order wins (a local skill can shadow a bundled one). Only load skills you trust, and see the skill trust model before loading external skills.

To list locally available skills, inspect path resolution, check load readiness, or run bundle tests from the terminal, see the CLI reference (skillware list, skillware paths, skillware doctor, skillware test).

Provider Adapter Guide Agent API key (typical)
Google Gemini to_gemini_tool() gemini.md GOOGLE_API_KEY (install skillware[gemini] for google-genai)
Anthropic Claude to_claude_tool() claude.md ANTHROPIC_API_KEY
OpenAI (ChatGPT) to_openai_tool() openai.md OPENAI_API_KEY
OpenAI-compatible hosts to_openai_tool() openai_compatible.md Host-specific key
DeepSeek to_deepseek_tool() deepseek.md DEEPSEEK_API_KEY
Ollama (prompt mode) to_ollama_prompt() ollama.md (local; no cloud key)
CLI skillware list, skillware paths, skillware doctor, skillware test, skillware examples cli.md pytest in [dev] for test
Install extras Category, skill, SDK, and meta pip install targets install_extras.md See guide for [all], [agents], per-skill extras

Skill-specific Usage Examples (sample prompts and execute payloads) live on each skill catalog page.

Shared patterns (load bundle, run execute, return tool results): agent_loops.md. After load_skill, prefer bundle["class"]() to instantiate the skill; explicit bundle["module"].ClassName() also works. Runnable script inventory: examples/README.md.

Contributors adding Usage Examples to skill catalog pages: skill_usage_template.md.

Skills that call external APIs during execution: API keys for skills.