Skip to main content

Agent Skill Management System with self-growth capabilities

Project description

Skill Manager

Agent Skill Management System with self-growth capabilities.

Features

  • Skill Storage: SQLite + sqlite-vec for local storage with vector search
  • Semantic Query: ANN retrieval + dependency expansion + token budget cropping
  • Self-Growth: Automatically extract knowledge from conversations and tasks
  • Human-in-the-Loop: Review queue for approving/rejecting draft skills
  • Permission Separation: Agent code can only write drafts, humans control canonical
  • CLI Tool: Command-line interface for skill management

Installation

pip install skill-manager

Quick Start

from skill_manager import SkillManager
from skill_manager.models import Skill, Chunk

# Initialize
sm = SkillManager("skills.db")

# Register a skill
skill = Skill(
    id="frontend-react",
    name="Frontend React",
    chunks=[
        Chunk(
            id="frontend-react::hooks",
            skill_id="frontend-react",
            section="React Hooks",
            content="## useEffect\n\nUse useEffect for side effects...",
            tokens=50,
        )
    ]
)
sm.register(skill)

# Query skills
context = sm.query("React hooks best practices", budget=2000)
print(context)

# Extract knowledge from conversation
draft_id = sm.extract(messages)

# Review queue
pending = sm.review_queue()

# Approve draft
sm.approve(draft_id)

CLI Usage

# List skills
skill-cli list

# Review queue
skill-cli review-queue

# Approve/reject
skill-cli approve draft::skill-id
skill-cli reject draft::skill-id --reason "Not useful"

# Export/Import
skill-cli export frontend-react > react.json
skill-cli import react.json

# Statistics
skill-cli stats

Architecture

Agent Code (Python)
    │
    ├── query()          → Read canonical skills
    ├── extract()        → Write draft skills
    │
Human (CLI)
    │
    ├── register()       → Write canonical skills
    ├── approve()        → Promote draft → canonical
    ├── reject()         → Archive draft
    └── delete()         → Remove skill
    │
SkillManager
    │
    ├── Storage (SQLite + sqlite-vec)
    ├── Query Engine (ANN + Dependencies + Budget)
    ├── Growth Engine (LLM extraction)
    └── Review System (Approve/Reject)

Development

# Install dependencies
pip install -e ".[dev]"

# Run tests
pytest

# Run CLI
skill-cli --help

License

MIT

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

skillvault-0.1.0.tar.gz (39.0 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

skillvault-0.1.0-py3-none-any.whl (18.6 kB view details)

Uploaded Python 3

File details

Details for the file skillvault-0.1.0.tar.gz.

File metadata

  • Download URL: skillvault-0.1.0.tar.gz
  • Upload date:
  • Size: 39.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.12

File hashes

Hashes for skillvault-0.1.0.tar.gz
Algorithm Hash digest
SHA256 3b476dd8b029f0102f037244d38165aeebcf4e4ac70692b43037363cf881dddc
MD5 e26eecf33cd07d362ead696e7ff0030e
BLAKE2b-256 dabc7b3329192772887b8b3296b7c0e4746cbaf49b2497e0fe61edfcd98bd1a9

See more details on using hashes here.

File details

Details for the file skillvault-0.1.0-py3-none-any.whl.

File metadata

  • Download URL: skillvault-0.1.0-py3-none-any.whl
  • Upload date:
  • Size: 18.6 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.12

File hashes

Hashes for skillvault-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 c4b5cf87bdb4dde92566b53084428c7d766b4d16fbd8d5e2a4219af86c8ec599
MD5 81db4e58ed143d557de0966c54bde781
BLAKE2b-256 962630918417214d3514b9925cff8c47cea2fe26777fc32abfcba4e4410a0cc8

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page