← Back to Skills Marketplace
mirni

Agentfinder

by mirni · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ Security Clean
81
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install gh-agentfinder
Description
Kayak for agents — search across ClawHub, SkillsMP, LobeHub, and more to find the right skill for your task. Compare results across registries, get recommend...
README (SKILL.md)

AgentFinder

Find the right skill across multiple registries in one call.

Start the server

uvicorn agentfinder.app:app --port 8016

Search across all registries

curl -s -X POST http://localhost:8016/v1/search \
  -H "Content-Type: application/json" \
  -d '{"query": "security scanning"}' | jq

Returns results sorted by relevance, from ClawHub, SkillsMP, LobeHub, and more.

Compare results by registry

curl -s http://localhost:8016/v1/search/security/compare | jq

Get recommendations for a problem

curl -s -X POST http://localhost:8016/v1/recommend \
  -H "Content-Type: application/json" \
  -d '{"problem": "I need to validate data before passing it to another agent"}' | jq

Filter by registry

curl -s -X POST http://localhost:8016/v1/search \
  -H "Content-Type: application/json" \
  -d '{"query": "code review", "registries": ["clawhub"], "max_results": 5}' | jq

List available registries

curl -s http://localhost:8016/v1/registries | jq

Endpoints

Method Path Description
POST /v1/search Search across registries
GET /v1/search/{query}/compare Compare by registry
POST /v1/recommend Recommend skills for a problem
GET /v1/registries List available registries
Usage Guidance
This skill appears internally consistent and limited to running a local FastAPI server that searches a built-in skill index. Before installing: 1) be aware it will install Python packages from PyPI (fastapi, uvicorn, pydantic) — install in a virtualenv or container if you prefer isolation; 2) the index is local/static, so it won't fetch live registry data unless modified to do so — review agentfinder/registries.py if you expect live scraping or syncing; 3) there is no network exfiltration or secret access in the current code, but if the owner adds adapters that call external registries you should review any new network code and required credentials. If you want, I can point out the exact lines that are safe or highlight where to add live-registry connectors safely.
Capability Analysis
Type: OpenClaw Skill Name: gh-agentfinder Version: 1.0.0 The gh-agentfinder skill bundle is a legitimate search aggregator for agent registries. It implements a FastAPI application that searches a hardcoded local index of skills using fuzzy matching. The code (app.py, registries.py) and instructions (SKILL.md) are well-structured, lack any dangerous execution patterns, and do not perform unauthorized network or file system operations.
Capability Tags
crypto
Capability Assessment
Purpose & Capability
Name/description (AgentFinder: aggregator/search across registries) match the code and SKILL.md. The package only needs Python and uses a local index of registries/skills; URLs in the index are references only. No unrelated credentials, binaries, or config paths are requested.
Instruction Scope
SKILL.md tells the operator to run uvicorn locally and to call local endpoints (search, compare, recommend, registries). The instructions do not ask the agent to read system files, environment secrets, or transmit data to external endpoints; runtime behavior is limited to the local API described.
Install Mechanism
Install spec indicates installing well-known Python packages (fastapi, uvicorn, pydantic) via pip. These are standard dependencies for a local FastAPI app. The install kind labeled 'uv' is uncommon but SKILL.md metadata shows pip package names; nothing is being downloaded from an untrusted custom URL or executed from a remote archive.
Credentials
The skill declares no required environment variables, no credentials, and no config paths. The code does not reference environment secrets. Requested privileges are minimal and proportional to a local search API.
Persistence & Privilege
always is false and the skill does not request permanent presence or modify other skills or system-wide settings. It exposes a local HTTP API only when the operator runs the server.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install gh-agentfinder
  3. After installation, invoke the skill by name or use /gh-agentfinder
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of gh-agentfinder: - Search and compare agent skills across ClawHub, SkillsMP, LobeHub, and more registries. - Get tailored agent recommendations for your task or problem. - Filter search results by specific registries with customizable result limits. - List all available registries for agent discovery. - Simple REST API with endpoints for search, comparison, recommendations, and registry listing.
Metadata
Slug gh-agentfinder
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is Agentfinder?

Kayak for agents — search across ClawHub, SkillsMP, LobeHub, and more to find the right skill for your task. Compare results across registries, get recommend... It is an AI Agent Skill for Claude Code / OpenClaw, with 81 downloads so far.

How do I install Agentfinder?

Run "/install gh-agentfinder" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Agentfinder free?

Yes, Agentfinder is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Agentfinder support?

Agentfinder is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Agentfinder?

It is built and maintained by mirni (@mirni); the current version is v1.0.0.

💬 Comments