Skip to content
Back to search

skill-router

MarketplaceWorks with Claude CodeGitHub

skill-router: picks the right skills, plugins and MCP servers for every Claude Code message.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/fripouilledu92--skill-router/badge.svg)](/listings/fripouilledu92--skill-router)

HTML

<a href="/listings/fripouilledu92--skill-router"><img src="/api/listings/fripouilledu92--skill-router/badge.svg" alt="findsafeskills safety rating" /></a>

Install

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

claude plugin marketplace add Fripouilledu92/skill-router
claude plugin install skill-router@skill-router

Open a new Claude Code session. Your tools are classified once in the background, then every message is routed locally in about 0.1 s, with no model call.

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 14 lines
  1. {
  2. "name": "skill-router",
  3. "description": "skill-router: picks the right skills, plugins and MCP servers for every Claude Code message.",
  4. "owner": {
  5. "name": "Fripouilledu92"
  6. },
  7. "plugins": [
  8. {
  9. "name": "skill-router",
  10. "source": "./",
  11. "description": "Picks the skills, plugins and connected MCP servers that fit each message, in the right order, and explains why."
  12. }
  13. ]
  14. }
README.md · README · 108 lines
  1. <p align="center"><img src=".claude-plugin/icon.png" width="96" alt="skill-router icon"></p>
  2. # skill-router
  3. **Your Claude Code agent has 100+ skills, plugins and MCP servers. It uses three.**
  4. skill-router picks the right ones for every message, hands them to Claude in the order a pro would use them,
  5. and Claude ends each answer with one line saying what each tool brought.
  6. <p align="center"><img src="assets/demo.png" alt="skill-router choosing brainstorming, a design skill, unslop, break-ui and deploy for a landing page request"></p>
  7. <p align="center"><sub>Illustration: the exact list depends on the tools you have installed.</sub></p>
  8. ## Install
  9. ```bash
  10. claude plugin marketplace add Fripouilledu92/skill-router
  11. claude plugin install skill-router@skill-router
  12. ```
  13. Open a new Claude Code session. Your tools are classified once in the background, then every message is routed
  14. locally in about 0.1 s, with no model call.
  15. ## How it works
  16. | When | What happens | Cost |
  17. |---|---|---|
  18. | Install / new session | Every skill, agent, command, plugin and **connected** MCP server is listed. New or changed ones are classified once, in the background, by Claude: category, tags, specialty, phase, neighbors. | One `claude -p` call per batch of new tools, once |
  19. | Every message | A local script (no model call) scores your message against the index and gives Claude a shortlist, the usual chain for that kind of task, and the nuances between similar tools. | 0 model calls, ~250-400 tokens of context, ~0.1 s |
  20. | Claude answers | Claude keeps the tools that really help, uses them in order (frame → design → build → verify → ship) and ends with a `Tools:` line. | — |
  21. | After the answer | It notes which proposed tools were actually used; useful ones climb, ignored ones sink a little. | 0 |
  22. **Neighbors, not duplicates.** Two design skills are not the same skill. Each tool carries a specialty and a
  23. "when to pick me over my neighbor" nuance, so Claude chooses `unslop` to clean an AI-looking screen and
  24. `impeccable` to design one from scratch.
  25. **Only what is connected.** MCP servers that need auth or failed to start are ignored (`claude mcp list`).
  26. **Quiet on small talk.** "ok", "thanks", "continue" and slash commands trigger nothing.
  27. ## Commands
  28. | Command | What it does |
  29. |---|---|
  30. | `/why` | Why these tools were proposed for the last message, which words matched, which were really used |
  31. | `/skills-map` | Your whole toolbox, by category, with security notes |
  32. | `/skills-off <name>` · `/skills-off on <name>` | Never propose a tool again / bring it back |
  33. | `/skills-sync` | Classify new tools now, including the ones only the agent can see (desktop app connectors) |
  34. ## Requirements
  35. Claude Code, Python 3.9+, the `claude` CLI on your PATH. The first classification of ~150 tools takes a few
  36. minutes in the background; until then tools are matched on their raw descriptions.
  37. ## Settings
  38. `~/.claude/skill-router/prefs.json`:
  39. ```json
  40. { "max": 5, "modele": "sonnet", "langue": "english", "classement_auto": true, "off": [] }
  41. ```
  42. ## Fixing a classification
  43. Anything the automatic classification got wrong goes in `~/.claude/skill-router/corrections.json`. Corrections
  44. always win over the classifier and survive re-classification:
  45. ```json
  46. {
  47. "categories": { "Old name": "New name" },
  48. "items": { "skill:unslop": { "phase": "verifier", "specialite": "…" } },
  49. "recettes": { "Design": [ { "quand": "…", "mots": ["site"], "chaine": ["skill:a", "skill:b|skill:c", "skill:d"] } ] }
  50. }
  51. ```
  52. `"skill:b|skill:c"` in a chain means "one of these, pick by the nuances".
  53. ## Security
  54. - Classification only **reads** tool files. It never runs a skill's scripts.
  55. - Each tool gets an `alerte` field when it sends data out, publishes, spends money or touches your config;
  56. `/skills-map` shows them.
  57. - The router only ever *suggests*. Your permission settings still apply to everything Claude does.
  58. ## Limits (v0.1)
  59. - Claude Code only. Codex, Cursor and Gemini CLI adapters are planned.
  60. - claude.ai on the web has no hooks, so it cannot run there automatically.
  61. - Matching is lexical (with FR/EN synonyms) for speed and zero cost: it builds a generous shortlist and lets
  62. Claude make the final call.
  63. - Learning ignores claude.ai connectors (their tool names are opaque ids).
  64. ## Tests
  65. The cases are examples from the author's own setup: they check the router against *your* index
  66. (`~/.claude/skill-router/index.json`), so adapt `tests/cas.json` to the tools you have.
  67. ```bash
  68. python3 tests/evaluer.py # labelled cases used while tuning
  69. python3 tests/evaluer.py tests/cas-inedits.json # cases written after tuning
  70. ```
  71. ---
  72. **En français** : à chaque message, skill-router choisit les skills, plugins et serveurs MCP connectés qui servent
  73. la demande, les donne à Claude dans l'ordre, et Claude termine sa réponse par une ligne « Outils : … ». Le tri se
  74. fait une fois à l'installation et à chaque ajout ; l'analyse de chaque message est locale et ne coûte aucun appel
  75. au modèle.
  76. MIT License.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.