Skip to content
All flagships

Alpha / active rewrite

Synthesis

Find and compose an existing skill before creating another.

A skill acquisition workflow for coding agents: search installed and curated packages, compose what is available, then create a local draft when something is missing. New drafts remain untrusted until review.

Open the alpha quickstart Follow the guide

What you can do with it

Who it's for: Developers experimenting with reusable skills and reviewable agent self-extension.

An inspectable skill acquisition result and a local package to review before use.

Your first useful result

  1. 01

    Install from the source checkout

    Follow the repository installation section. Run this inside the cloned Synthesis repository.

    pip install -e ".[dev]"
  2. 02

    Try the documented mock workflow

    Save this as a Python script in your test directory. It requests a CSV skill and prints the acquisition result.

    import asyncio
from synthesis import SynthesisClient

async def main():
    client = SynthesisClient(provider_type="mock", host_root="./installed-skills")
    result = await client.acquire_skill("parse csv files")
    print(result.to_dict())

asyncio.run(main())
  3. 03

    Inspect the result before using it

    Check the returned source and lifecycle status. Read the skill package and its validation output before connecting it to an agent with access to your project.

Know the limits

  • The current alpha is being rewritten around GitHub skill curation. The older federated marketplace paper describes a broader design.
  • A passing test suite or a mock acquisition result does not prove a generated skill is safe or useful.

Source & further reading