cli-skill-collector
Skill0 starsChanged 3 months agoMITView on GitHub
Detect installed CLI coding tools (Claude Code, Codex, Cursor, Copilot, Cline and more), search GitHub for matching agent skills, and install them to the detected tools via Nerve's built-in APIs.
- Reads credentials
- Installs packages
- Sends data out
3 other projects use a version of this, 3 of them edited.
Adopted by
Most recently taken up by
- MMACABRE/SKILLSedited, first seen Nov 8, 2026
- cheaper-inference/OmniRouteedited, first seen Oct 25, 2026
- diegosouzapw/OmniRouteedited, first seen Oct 22, 2026
Other versions3 projects use a version of this.
Word for word · 0
None found.
Edited · 3
- cheaper-inference/OmniRoute★ 86k83% same
- diegosouzapw/OmniRoute★ 86k83% same
- MMACABRE/SKILLS★ 184% same
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository49 other files it gives its agents.
- AGENTS.mdAGENTS.md
- app_backup/AGENTS.mdAGENTS.md
- app_backup/CLAUDE.mdCLAUDE.md
- CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- cli-a2aSkill
- cli-backup-syncSkill
- cli-batchesSkill
- cli-chatSkill
- cli-compressionSkill
- cli-contextsSkill
- cli-cost-usageSkill
- cli-evalSkill
- cli-healthSkill
- cli-keysSkill
- cli-mcpSkill
- cli-modelsSkill
- cli-plugins-skillsSkill
- cli-policy-auditSkill
- cli-providersSkill
- cli-resilienceSkill
- cli-routingSkill
- cli-serveSkill
- cli-setupSkill
- cli-tunnelSkill
- config-codex-cliSkill
- nerve-inferenceSkill
- nerve-mcpSkill
- omni-agents-a2aSkill
- omni-api-keysSkill
- omni-authSkill
- omni-budgetSkill
- omni-cacheSkill
- omni-cli-toolsSkill
- omni-combos-routingSkill
- omni-compressionSkill
- omni-context-rtkSkill
- omni-db-backupsSkill
- omni-github-skillsSkill
- omni-modelsSkill
- omni-providersSkill
- omni-proxiesSkill
- omni-resilienceSkill
- omni-settingsSkill
- omni-sync-cloudSkill
- omni-tunnelsSkill
- omni-usage-logsSkill
- omni-version-managerSkill
- omni-webhooksSkill
The file
We split this skill into the rules it gives an agent
4 rules. 2 are common, 0 are distinctive, and 2 are very specific to this project.
2
Common
Most of the 3 edited versions we read keep these.
0
Distinctive
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Common: line 100
Write config for a tool
- Very specific: line 284
No separate Skill Collector app needed — Nerve's own CLI detection + GitHub search handles everything.
- Very specific: line 298
Every CLI tool in Nerve's catalog (CLI_TOOL_IDS: claude, codex, cursor, copilot, opencode, cline, kilocode, hermes, hermes-agent, openclaw, …
- Common: line 371
After installing, verify the skill is in place:
“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
<!-- generated by src/lib/agentSkills/generator.ts; manual edits will be overwritten -->Overview
Detect installed CLI coding tools (Claude Code, Codex, Cursor, Copilot, Cline and more), search GitHub for matching agent skills, and install them to the detected tools via Nerve's built-in APIs.
Quick install
npm install -g nerve # or: npx nerve
nerve --versionSubcommands
autostart
Example:
nerve autostartautostart enable
Example:
nerve autostart enableautostart disable
Example:
nerve autostart disableautostart toggle
Example:
nerve autostart toggleautostart status
Example:
nerve autostart statusconfig
Show or update CLI tool configuration
Example:
nerve configconfig list
List all CLI tools and config status
Flags:
--json
Example:
nerve config listconfig get <tool>
Show current config for a tool
Flags:
--json
Example:
nerve config get <tool>config set <tool>
Write config for a tool
Flags:
--model <model>--non-interactive--yes
Example:
nerve config set <tool>config validate <tool>
Validate config format without writing
Flags:
--model <model>--json
Example:
nerve config validate <tool>config opencode
Generate OpenCode config (alias for
Flags:
--model <model>--non-interactive--yes
Example:
nerve config opencodeconfig lang
Example:
nerve config langconfig get
Flags:
--json
Example:
nerve config getconfig set <code>
Flags:
--force
Example:
nerve config set <code>config list
Flags:
--json
Example:
nerve config listenv
Show and manage environment variables
Example:
nerve envenv show
Show current environment variables
Flags:
--json
Example:
nerve env showenv get <key>
Get a single environment variable
Example:
nerve env get <key>env set <key> <value>
Set an environment variable (current session only)
Example:
nerve env set <key> <value>setup
Flags:
--password <value>--add-provider--provider <id>--provider-name <name>--api-key <value>--default-model <model>--provider-base-url <url>--test-provider--non-interactive--list
Example:
nerve setupupdate
Flags:
--check--apply--changelog--dry-run--no-backup--yes
Example:
nerve update<!-- skill:custom-start --><!-- Preserved curated content from #6294 (skills/cli-skill-collector authored workflow) -->/cli-skill-collector — Agent Skill Collector
Discover and install agent skills for your coding CLI tools — all through Nerve's built-in APIs.
This skill teaches you how to:
- Detect which coding CLIs are installed on this machine
- Search GitHub for relevant agent skills (SKILL.md repos)
- Install discovered skills to the detected coding tools
No separate Skill Collector app needed — Nerve's own CLI detection + GitHub search handles everything.
Step 1 — Detect installed coding tools
Query Nerve's CLI tool detection to find which coding agents are installed:
curl -H "Authorization: Bearer $NERVE_API_KEY" http://localhost:20128/api/skills/collect/detectThis returns:
- Every CLI tool in Nerve's catalog (
CLI_TOOL_IDS: claude, codex, cursor, copilot, opencode, cline, kilocode, hermes, hermes-agent, openclaw, droid, continue, qwen, windsurf, devin, antigravity, etc.) - Whether each is installed and runnable
- GitHub skills matched to your installed tools (scored by relevance)
Example response:
{
"tools": {
"codex": { "installed": true, "runnable": true, "command": "codex" },
"claude": { "installed": true, "runnable": true, "command": "claude" },
"cursor": { "installed": false, "runnable": false }
},
"installedToolIds": ["codex", "claude"],
"matchedSkills": [
{ "toolId": "codex", "repo": "user/skill-codex-xxx", "score": 0.85, "stars": 120 },
{ "toolId": "claude", "repo": "user/claude-agent-rules", "score": 0.92, "stars": 340 }
],
"totalSkills": 85
}Step 2 — Review matched skills
For each installed tool, the API returns relevant GitHub repos that contain SKILL.md or agent configuration files. Use the score field to prioritize:
| Score | Recommendation |
| 0.80+ | Excellent — well-maintained, high stars, active |
| 0.60+ | Good — relevant with decent quality |
| 0.40+ | Fair — may need review |
| <0.40 | Low quality — skip |
You can also browse manually:
curl -H "Authorization: Bearer $NERVE_API_KEY" \
"http://localhost:20128/api/github-skills?minStars=3&maxResults=50"Step 3 — Install skills to detected tools
Install a chosen skill to one or more detected tools:
curl -X POST http://localhost:20128/api/skills/collect/install \
-H "Authorization: Bearer $NERVE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"repoName": "user/skill-codex-xxx",
"targets": ["codex", "claude"],
"description": "Agent skill for coding workflows"
}'This plans the installation path for each target tool:
- claude →
~/.claude/skills/{category}/ - codex →
~/.codex/skills/{category}/ - hermes →
~/AppData/Local/hermes/skills/{category}/ - opencode →
~/.opencode/skills/{category}/ - gemini →
~/.gemini/skills/{category}/
The actual file sync (cloning from GitHub and copying SKILL.md) is done by the agent using standard curl + cp commands.
Step 4 — Verify installation
After installing, verify the skill is in place:
# For Codex
ls -la ~/.codex/skills/imported-github/*/SKILL.md
# For Claude Code
ls -la ~/.claude/skills/imported-github/*/SKILL.md
# For Hermes (Windows)
ls -la ~/AppData/Local/hermes/skills/imported-github/*/SKILL.mdAlso re-check detection:
curl -H "Authorization: Bearer $NERVE_API_KEY" http://localhost:20128/api/skills/collect/detectQuick start (full workflow)
AUTH_HEADER="Authorization: Bearer $NERVE_API_KEY"
# 1. Detect
DETECT=$(curl -s -H "$AUTH_HEADER" http://localhost:20128/api/skills/collect/detect)
# 2. Pick top matched skill for first installed tool
TOOL=$(echo "$DETECT" | python3 -c "import sys,json;d=json.load(sys.stdin);print(d['installedToolIds'][0] if d['installedToolIds'] else '')")
SKILL=$(echo "$DETECT" | python3 -c "import sys,json;d=json.load(sys.stdin);ms=d.get('matchedSkills',[]);print(ms[0]['repo'] if ms else '')")
if [ -n "$TOOL" ] && [ -n "$SKILL" ]; then
# 3. Install
curl -s -X POST http://localhost:20128/api/skills/collect/install \
-H "$AUTH_HEADER" \
-H "Content-Type: application/json" \
-d "{\"repoName\": \"$SKILL\", \"targets\": [\"$TOOL\"]}"
echo "Installed $SKILL to $TOOL"
fiNotes
- Nerve must be running locally on port 20128 (default) — see
docs/frameworks/SKILLS.mdfor custom-port setups. - The
/api/skills/collect/*and/api/github-skillsendpoints require management-scoped authentication the same way every other/api/skills/*route does: a dashboard session, the loopback CLI token, or an API key with themanagescope (requireManagementAuth()). Auth is only bypassed when the server has no login/API-key requirement configured at all. - This replaces the standalone Skill Collector Python app — all logic is now inside Nerve.
<!-- skill:custom-end -->Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

