Python API Reference¶
The skills_tree Python package provides a clean API for programmatic access to the skill taxonomy.
Installation¶
SkillsTree¶
The main entry point.
SkillsTree.get(skill_id: str) -> Skill¶
Fetch a skill by its ID.
skill = st.get("rag")
print(skill.title) # "Retrieval-Augmented Generation"
print(skill.category) # "memory"
print(skill.version) # "v3"
print(skill.badge) # "verified"
Raises: SkillNotFound if the skill ID does not exist.
SkillsTree.search(query: str, limit: int = 20) -> list[Skill]¶
Full-text search across all skill titles, descriptions, and tags.
results = st.search("memory injection", limit=5)
for skill in results:
print(skill.id, skill.title, skill.badge)
SkillsTree.categories() -> list[Category]¶
List all 17 skill categories.
SkillsTree.get_category(category_id: str) -> list[Skill]¶
Get all skills in a specific category.
SkillsTree.recommend(task: str, top_k: int = 5) -> list[Skill]¶
Recommend skills for a given task description using semantic similarity.
recommendations = st.recommend(
"I need to build an agent that remembers user preferences",
top_k=5
)
Data Models¶
Skill¶
| Field | Type | Description |
|---|---|---|
id | str | Unique skill identifier (slug) |
title | str | Human-readable skill name |
category | str | Parent category ID |
level | str | beginner, intermediate, or advanced |
stability | str | experimental, beta, or stable |
version | str | Current version (e.g. v3) |
badge | str | verified, reviewed, or stub |
tags | list[str] | Associated tags |
related | list[str] | Related skill IDs |
content | str | Full Markdown content |
Category¶
| Field | Type | Description |
|---|---|---|
id | str | Category identifier (e.g. memory) |
name | str | Display name |
skill_count | int | Number of skills in this category |