Route a coding or design task through SkillRadar's bundled safety-gated registry.

MITAuto-check passed

Install Skill Router

skills CLI
$ npx skills add changchangidea-oss/SkillRadar --skill skill-router -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install changchangidea-oss/SkillRadar skill-router --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/changchangidea-oss/SkillRadar.git skills-src && mkdir -p .claude/skills && cp -r skills-src/packages/codex-plugin/skills/skill-router .claude/skills/skill-router && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
skill-router
GitHub stars
190
Token cost
~1.5k tokens
SKILL.md length
740 words
Files
1
Skills in repo
5
Repo updated
First seen
Licence
MIT

At a glance

Route a coding or design task through SkillRadar's bundled safety-gated registry.

  • Works in 10 steps: Convert the user's task into a concise… → For explicit $skill-router,… → Read the canonical registry evidence:… → …
  • The user explicitly invokes $skill-router
  • SKILL.md covers Mandatory routing contract, Trusted internal lookup, Matching v2.1 and Strong recommendation contract, plus 3 more sections
  • Calls node and git

What it does

Skill Router is an agent skill from changchangidea-oss/SkillRadar. Route a coding or design task through SkillRadar's bundled safety-gated registry. Use when the user explicitly invokes $skill-router, asks for Top N skills, asks which skill to use, requests skill recommendations/comparison, or when selecting a better skill before implementation would materially improve the result.

Its SKILL.md is about 1.5k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

The licence is MIT.

When your agent uses it

  • The user explicitly invokes $skill-router
  • Asks for Top N skills
  • Asks which skill to use
  • Requests skill recommendations/comparison

Example prompts

  • “/skill-router”

Workflow steps

10 steps, taken from the first numbered list in SKILL.md.

  1. Convert the user's task into a concise routing query without dropping important frameworks, technologies, design disciplines, or requested…
  2. For explicit $skill-router, recommendation, comparison, ranking, or Top N requests, run node ../../scripts/skillradar.mjs match '' before…
  3. Read the canonical registry evidence: top-level source, registry.mode, ranking.version, context.mode, capability_gap, and for each result…
  4. Treat match_details.matched_signals / coverage as task evidence. Treat project_context_signals / project_context_bonus only as secondary…
  5. Return up to 3 strong Registry recommendations by default unless the user asks for another N. Preserve SkillRadar ordering unless you…
  6. Prefer A/B security skills when relevance is close. If the Top-1 result is C and SkillRadar returns an advisory, surface the safer nearby…
  7. Only after the Registry result is known, inspect whether those candidates are already installed locally. Local availability is…
  8. If a Registry candidate is not installed, show its source/install URL before running any third-party scripts. Do not silently swap it for…
  9. For design requests, preserve the discipline intent: UI, visual communication, operations, video, industrial, environmental, fashion…
  10. Return to the original task after selecting the capability.

What it can do on your machine

Read from SKILL.md and the folder at commit be18131. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • node
    • git

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use git, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Skill Router loads about 1.5k tokens when it runs. Until then it costs about 82 tokens; SKILL.md has 740 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~82
When it runs · the whole SKILL.md, loaded when a task matches
~1.5k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from changchangidea-oss/SkillRadar at commit be18131, republished under its MIT licence (© changchangidea-oss). 740 words, ~1,482 tokens.

Download SKILL.mdSave it as .claude/skills/skill-router/SKILL.md (or your agent's skills folder).
name
skill-router
description
Route a coding or design task through SkillRadar's bundled safety-gated registry. Use when the user explicitly invokes $skill-router, asks for Top N skills, asks which skill to use, requests skill recommendations/comparison, or when selecting a better skill before implementation would materially improve the result.

Skill Router

Use SkillRadar as a retrieval layer, not as a reason to load every skill into context.

Mandatory routing contract

When this skill is explicitly invoked with $skill-router, or the user asks for skill recommendations, comparison, ranking, routing, or a Top N list, you MUST query SkillRadar's bundled registry before choosing skills.

Do not satisfy an explicit routing request only by inspecting Codex's currently loaded or installed skills. Installed skills may be noted after routing as an availability signal, but they must not replace the SkillRadar registry result.

For explicit routing requests, the first routing command must be:

node ../../scripts/skillradar.mjs match '<user task>'

Do not use web search, GitHub search, git clone, or ad-hoc local skill inspection as a substitute when this command succeeds.

Trusted internal lookup

node ../../scripts/skillradar.mjs ... is the read-only lookup CLI shipped inside the installed SkillRadar plugin. It reads SkillRadar's bundled safety-gated registry and does not install or execute any discovered third-party skill. Treat this internal lookup differently from scripts that belong to candidate third-party skills.

Matching v2.1

SkillRadar v0.5 keeps task evidence primary, then uses a bounded project-context signal as a secondary tie-break. The CLI may read non-secret project metadata such as package.json, common config filenames, and framework directories in the current working directory. It does not read source files, environment-variable contents, credentials, or third-party scripts for project context.

Project context can add only a small bounded bonus and must be reported separately from task evidence. A currently installed framework must not replace or reinterpret the user's explicit task.

Strong recommendation contract

match returns up to 3 strong recommendations, not an artificially padded Top 3. SkillRadar's complementary reranker already applies a relative strong-match floor. Candidates below that floor must not be backfilled only to reach three results.

When fewer than three candidates meet the floor, the CLI returns top-level capability_gap metadata with detected: true, the number returned, and the number missing. Treat that as useful evidence that the current Registry is sparse for this task. Do not hide the gap by substituting unrelated local skills, arbitrary web results, or manually invented candidates.

A capability gap is a discovery/coverage signal, not permission to install anything automatically. It should feed later Radar discovery and evaluation work.

Show full SKILL.md (371 more words)Show less

Workflow

  1. Convert the user's task into a concise routing query without dropping important frameworks, technologies, design disciplines, or requested capabilities.
  2. For explicit $skill-router, recommendation, comparison, ranking, or Top N requests, run node ../../scripts/skillradar.mjs match '<task>' before selecting candidates.
  3. Read the canonical registry evidence: top-level source, registry.mode, ranking.version, context.mode, capability_gap, and for each result match_score, skillradar_score, security, source, reason, and match_details.
  4. Treat match_details.matched_signals / coverage as task evidence. Treat project_context_signals / project_context_bonus only as secondary context evidence.
  5. Return up to 3 strong Registry recommendations by default unless the user asks for another N. Preserve SkillRadar ordering unless you clearly explain a safety override. If capability_gap.detected is true, surface the gap instead of padding the list.
  6. Prefer A/B security skills when relevance is close. If the Top-1 result is C and SkillRadar returns an advisory, surface the safer nearby alternative.
  7. Only after the Registry result is known, inspect whether those candidates are already installed locally. Local availability is supplemental metadata, not a replacement ranking.
  8. If a Registry candidate is not installed, show its source/install URL before running any third-party scripts. Do not silently swap it for a different installed skill.
  9. For design requests, preserve the discipline intent: UI, visual communication, operations, video, industrial, environmental, fashion, experience, digital media/film, arts & crafts, folk art, or architecture.
  10. Return to the original task after selecting the capability.

Required response evidence

For explicit routing/recommendation requests, include enough evidence to prove the bundled Registry and matching v2.1 were used:

  • source: skillradar-registry
  • registry.mode: local-bundled when the bundled snapshot is healthy
  • ranking.version: 2.1
  • context.mode: project-aware or task-only
  • capability_gap when fewer than 3 strong recommendations are returned
  • for every recommended result: match_score, skillradar_score, security, source, and reason
  • summarize useful match_details, especially task matched signals / coverage and any project-context bonus, when the ranking needs explanation

If these fields are unavailable because the lookup failed, say so explicitly. Do not present a local-skill-only shortlist as a successful SkillRadar route.

Safety

Discovery is not execution. Running SkillRadar's own read-only registry CLI is allowed for routing; it does not authorize installation or execution of matched third-party skills. Popularity is not permission. Third-party shell commands, secret access, database writes and deployments remain subject to normal authorization and safety checks.

© changchangidea-oss, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in packages/codex-plugin/skills/skill-router of changchangidea-oss/SkillRadar.

Open the folder on GitHubat commit be18131

Compare with similar skills

Skill Router next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Skill Router compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Skill Router this skillchangchangidea-oss/SkillRadar190—~1.5kAutomated safety check: PassMIT
Router Onweave-os/router5.6k—~169Automated safety check: PassApache-2.0
Acp Routeropenclaw/openclaw392k—~2.4kAutomated safety check: PassMIT
Router Offweave-os/router5.6k—~183Automated safety check: PassApache-2.0
OmniRoute Routing CLIdiegosouzapw/OmniRoute74k—~342Automated safety check: PassMIT
OmniRoute Combo Routingdiegosouzapw/OmniRoute74k—~2.1kAutomated safety check: PassMIT

Similar skills

  • Router On

    weave-os/router

    Route Codex through the Weave Router again (turn it back on).

    5.6k GitHub stars~169 tokensUpdated today
    Auto-check passed
  • Acp Router

    openclaw/openclaw

    Route plain-language requests for Claude Code, Cursor, Copilot, OpenClaw ACP, OpenCode, Gemini CLI, Qwen, Kiro, Kimi, iFlow, Factory Droid, Kilocode, or explicit ACP harness work into either…

    392k GitHub stars~2.4k tokensUpdated today
    Writing & ContentAuto-check passed
  • Router Off

    weave-os/router

    Route Codex to its default provider again (turn the Weave Router off).

    5.6k GitHub stars~183 tokensUpdated today
    Auto-check passed
  • OmniRoute Routing CLI

    diegosouzapw/OmniRoute

    Creates, switches, and inspects OmniRoute model-routing combos, plus a suggestion command with cost and latency constraints.

    74k GitHub stars~342 tokensUpdated today
    AI & LLM EngineeringAuto-check passed
  • OmniRoute Combo Routing

    diegosouzapw/OmniRoute

    Manages OmniRoute routing combos through its REST API: create and update combos, choose from 19 strategies, set fallback chains, test outcomes and read metrics.

    74k GitHub stars~2.1k tokensUpdated today
    Backend & APIsAuto-check passed
  • Explains LobeHub's split between src/routes page segments and src/features domain code, and where router config, redirects and platform adapters belong.

    83k GitHub stars~3.2k tokensUpdated today
    Frontend & DesignAuto-check passed

More from changchangidea-oss/SkillRadar

  • Skillradar

    changchangidea-oss/SkillRadar

    Discover, inspect, safety-check, and rank open Agent Skills before installing them.

    190 GitHub stars~732 tokensUpdated yesterday
    Auto-check passed
  • Manage Skills

    changchangidea-oss/SkillRadar

    Audit Codex skill-context pressure, detect duplicate capabilities, and produce a read-only pruning plan for global, project and plugin-provided skills.

    190 GitHub stars~614 tokensUpdated yesterday
    Auto-check passed
  • Find Skill

    changchangidea-oss/SkillRadar

    Search SkillRadar for agent skills by task, technology, design field, category, or tool.

    190 GitHub stars~229 tokensUpdated yesterday
    Auto-check passed
  • Inspect Skill

    changchangidea-oss/SkillRadar

    Inspect one SkillRadar skill for purpose, provenance, quality signals, security grade, maintenance, and suitability before installation or use.

    190 GitHub stars~216 tokensUpdated yesterday
    Auto-check passed

Questions about Skill Router

What does Skill Router do?

Route a coding or design task through SkillRadar's bundled safety-gated registry. Skill Router is an agent skill from changchangidea-oss/SkillRadar. Route a coding or design task through SkillRadar's bundled safety-gated registry.

When should I use Skill Router?

Skill Router fits situations like: the user explicitly invokes $skill-router; asks for Top N skills; asks which skill to use; requests skill recommendations/comparison.

How do I install Skill Router in Claude Code?

Run `npx skills add changchangidea-oss/SkillRadar --skill skill-router -a claude-code`. Or copy the skill folder (packages/codex-plugin/skills/skill-router in changchangidea-oss/SkillRadar) into .claude/skills/skill-router in your project. Claude Code loads it when a task matches its description.

How do I install Skill Router in Codex?

Run `npx skills add changchangidea-oss/SkillRadar --skill skill-router -a codex`. Or copy the skill folder (packages/codex-plugin/skills/skill-router in changchangidea-oss/SkillRadar) into .agents/skills/skill-router in your project. Codex loads it when a task matches its description.

Can I use Skill Router in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add changchangidea-oss/SkillRadar --skill skill-router -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/skill-router, .gemini/skills/skill-router, .github/skills/skill-router and .opencode/skills/skill-router in your project.

What does Skill Router need to run?

Going by SKILL.md and its folder, Skill Router needs the command-line tools its instructions call (node and git).

Does Skill Router access the network?

SKILL.md contains no URLs. Its commands use git, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Skill Router safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Skill Router use?

Skill Router is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Skill Router use?

About 1.5k tokens (SKILL.md is roughly 5.9k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Skill Router?

Skills that share tags, products or a category with Skill Router: Router On (weave-os/router, 5.6k stars), Acp Router (openclaw/openclaw, 392k stars), Router Off (weave-os/router, 5.6k stars) and OmniRoute Routing CLI (diegosouzapw/OmniRoute, 74k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Skill Router?

changchangidea-oss (a GitHub user) maintains it in changchangidea-oss/SkillRadar, which has 190 GitHub stars. The repository holds 5 skills in this directory. The repository was last updated on October 8, 2026.

Source: changchangidea-oss/SkillRadar on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.