Skip to main content

Skill and agent recommendation system for Claude Code — knowledge graph, wiki, and intake quality gates

Project description

ctx — Skill, Agent & MCP Recommendation for Claude Code

License: MIT Python 3.11+ PyPI Tests Graph Docs

Watches what you develop, walks a knowledge graph of 1,968 skills, 464 agents, and 10,786 MCP servers, and recommends the right bundle on the fly. You approve, it installs; you stop using, it unloads. Powered by a Karpathy LLM wiki with persistent memory that gets smarter every session.

2026-04-27 updates.

Why it exists

  • Discovery — with 1,700+ skills, 400+ agents, and 10K+ MCPs, you can't possibly know which exist or which apply to your current work.
  • Context budget — loading everything wastes tokens and degrades quality. You need the right 10–15 per session.
  • Skill rot — skills you installed months ago and never used are cluttering context. Stale ones should be flagged automatically.

Install

pip install claude-ctx
ctx-init --hooks            # one-shot setup: directories, hooks, starter toolboxes

Optional extras: pip install "claude-ctx[embeddings]" for the semantic backend, pip install "claude-ctx[dev]" for the test toolchain.

Pre-built knowledge graph (optional)

A pre-built knowledge graph of 13,218 nodes and 963K edges ships as a tarball. Extract to get a ready-to-use ~/.claude/skill-wiki/:

# after `git clone` — or download graph/wiki-graph.tar.gz from the GitHub release
mkdir -p ~/.claude/skill-wiki
tar xzf graph/wiki-graph.tar.gz -C ~/.claude/skill-wiki/

Windows / Git-Bash / MSYS: pass --force-local so tar doesn't read the c: in the path as a remote host: tar --force-local xzf graph/wiki-graph.tar.gz -C ~/.claude/skill-wiki/. Linux/macOS users can ignore.

Use

After install, the ctx hooks integrate automatically with Claude Code's PostToolUse + Stop events. Typical flow:

ctx-scan-repo --repo .     # scan current repo, surface recommended skills/agents
ctx-skill-quality list     # four-signal quality score for every skill
ctx-skill-quality explain python-patterns   # drill into a single skill
ctx-skill-health dashboard # structural health + drift detection
ctx-toolbox run --event pre-commit          # run a council on the current diff
ctx-monitor serve          # local dashboard: http://127.0.0.1:8765/

The ctx-monitor dashboard shows currently loaded skills with load/unload buttons, a cytoscape graph view (/graph?slug=…), the LLM-wiki entity browser (/wiki/<slug>), a filterable skills grid, a session timeline, an audit log viewer, and a live SSE event stream.

Full docs, architecture, and every module: https://stevesolun.github.io/ctx/

License

MIT — see LICENSE.

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

claude_ctx-0.7.0.tar.gz (439.2 kB view details)

Uploaded Source

Built Distribution

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

claude_ctx-0.7.0-py3-none-any.whl (505.2 kB view details)

Uploaded Python 3

File details

Details for the file claude_ctx-0.7.0.tar.gz.

File metadata

  • Download URL: claude_ctx-0.7.0.tar.gz
  • Upload date:
  • Size: 439.2 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for claude_ctx-0.7.0.tar.gz
Algorithm Hash digest
SHA256 e0da6733360534698ad5dbfcc88144ba61bff59893a8fc3c2d7e3502a271f718
MD5 4b06fb3f862fb056ae1e0d1a65363799
BLAKE2b-256 58807f7eedad30e0367921e135e623cb9c93a33370f919faba852b365d96b9d1

See more details on using hashes here.

Provenance

The following attestation bundles were made for claude_ctx-0.7.0.tar.gz:

Publisher: publish.yml on stevesolun/ctx

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file claude_ctx-0.7.0-py3-none-any.whl.

File metadata

  • Download URL: claude_ctx-0.7.0-py3-none-any.whl
  • Upload date:
  • Size: 505.2 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for claude_ctx-0.7.0-py3-none-any.whl
Algorithm Hash digest
SHA256 8f66f0b9a1b61eba904c9dc1172c14494affaf249cb705a7e15a795dbbaf1b5c
MD5 83d53abeb9593dbde4e598f17c84a53f
BLAKE2b-256 15b4659d65a4d9f22a1aee0b5c7428bac3186949e4d23c789966829b98bb758e

See more details on using hashes here.

Provenance

The following attestation bundles were made for claude_ctx-0.7.0-py3-none-any.whl:

Publisher: publish.yml on stevesolun/ctx

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

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