π open-source Β· MIT Β· v0.5.1 on PyPI
Trajectories in.
Trajectories in.
Reusable skills out.
One solves it. Everyone gets it.
Your coding agent re-derives the same solution every time it hits a familiar problem. xskill distills the patterns that actually worked into Skill files your agent loads automatically.
MIT-licensed Β· Python 3.9+ Β· no model weights touched.
π
SKILL.md
canary β
example β a Skill is a versioned git dir, kept by its UX score
π·οΈv0.5.1current release
π΄0agents supported
π3.9+Python
πMITopen source
how it works
A few narrow agents, one self-growing library
Every Skill is its own git repository β every change versioned and reversible.
- π
1 Watch
- πͺπ₯₯
2 Split into Atoms
- π§
3 Route to a Skill
- βοΈ
4 Rewrite SKILL.md
- π€
5 Canary A/B + UX score
works with your agents
Plug into the agent you already use
Trajectory ingest and Skill install are wired up per ecosystem.
| Agent | Status | Trajectory ingest | Skill install |
|---|---|---|---|
| π΄ Claude Code | verified | ~/.claude/projects/ | ~/.claude/skills/<name>/ |
| πΆ Codex CLI | verified | ~/.codex/sessions/ | ~/.agents/skills/<name>/ |
| π OpenCode | verified | opencode.db (SQLite) | ~/.agents/skills/<name>/ |
| π OpenClaw | not well tested | ~/.openclaw/agents/ | ~/.agents/skills/<name>/ |
| π Cursor | not well tested | ~/.cursor/.../agent-transcripts/ | ~/.cursor/skills/<name>/ |
| Any other agent | manual | xskill.adapters.submit_trajectory | copy / symlink SKILL.md |
Status as of v0.5.x Β· π‘ implemented but not well tested yet.
vocabulary
Five words and you've got it
Atom
Skill
Canary
UX score
the killer use case
Team mode: one library, the whole team
One machine is the server; everyone else joins as a thin client.
# Start a server, then join from any machine:
$ xskill serve --server
$ xskill connect <host:port> --token <token>
πͺ
Silently distill your top performers
π
Any workflow plugs in
π
Trajectories stay private
π§ͺ
A/B-driven evolution
π§βπ«
Experts can teach by hand
why xskill
Stop hand-maintaining a prompt library that rots
π₯₯
Distilled automatically
π±
Grows itself
β
Learns from your edits
β
UX-driven evolution
π
Distill your first Skill in two lines.
# install β Python 3.9+
$ pip install xskill
# writes ~/.xskill/config.yaml, then exits β add your model keys
$ xskill serve
# run again: auto-detects your agents and starts watching
$ xskill serve