AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
MCP verified MIT Self-run

Agent Tools

mcp-jesut0ni-agent-tools · by Jesut0ni

Open registry of MCP-callable tools for AI agents

No reviews yet
0 installs
40 views
0.0% view→install

Install

$ agentstack add mcp-jesut0ni-agent-tools

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access Used
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets Used
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/mcp-jesut0ni-agent-tools)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
2mo ago

Declared compatibility

Claude CodeClaude DesktopCursorWindsurf

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Agent Tools? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

agent-tools

An open registry of machine-readable tools designed for AI agents to discover and call programmatically; no human in the loop.

[](https://github.com/Jesut0ni/agent-tools/actions/workflows/ci.yml) [](LICENSE) [](CONTRIBUTING.md)

Built around three primitives:

  • Tool — a named, versioned, agent-callable capability (e.g. acme-send-invoice).
  • ToolSpec — the machine-readable contract: input schema, output schema, endpoint, auth requirements.
  • Invoke — a single POST /tools/:slug/call endpoint that proxies the call to the upstream endpoint with the agent's credentials.

Companion to AgentGateagent-tools is what gets called, AgentGate is the auth layer that decides who can call what.

Every tool registered here is also a native MCP tool, exposed over HTTP at /api/v1/mcp and over stdio via @agent-tools/mcp.

Status

v0.6 — auth, ownership, JSON Schema validation, real SSRF guard, rate limiting, CORS lockdown, AgentGate JWT integration, MCP server (HTTP + stdio), OpenAPI importer, 15 passing tests.

Stack

Hono · Drizzle · SQLite · TypeScript · Turbo workspaces · Zod · Ajv · jose · Next.js 15 · Tailwind.

Quickstart (dev)

npm install
npm run db:push
npm run seed       # creates a demo developer + 5 demo tools, prints the API key once
npm run dev        # api on :3002, web on :3000

Open http://localhost:3000. Sign up at /signup to get your own API key (or use the demo key printed by npm run seed).

Publish a tool (curl)

curl -X POST http://localhost:3002/api/v1/tools \
  -H 'Content-Type: application/json' \
  -H "Authorization: Bearer $AGENT_TOOLS_KEY" \
  -d '{
    "slug": "demo-echo",
    "name": "Echo",
    "description": "Echoes whatever you send it.",
    "version": "0.1.0",
    "spec": {
      "endpoint": { "method": "POST", "url": "https://postman-echo.com/post" },
      "input": { "type": "object", "properties": { "message": { "type": "string" } }, "required": ["message"] },
      "output": { "type": "object" },
      "auth": { "type": "none" }
    }
  }'

Invoke a tool

curl -X POST http://localhost:3002/api/v1/tools/demo-echo/call \
  -H 'Content-Type: application/json' \
  -d '{ "message": "hello agents" }'

Use it as an MCP server

HTTP transport:

curl -X POST http://localhost:3002/api/v1/mcp \
  -H 'Content-Type: application/json' \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'

Stdio bridge (for Claude Desktop / Cursor — see packages/mcp/README.md):

{
  "mcpServers": {
    "agent-tools": {
      "command": "npx",
      "args": ["-y", "@agent-tools/mcp"],
      "env": { "AGENT_TOOLS_API_URL": "http://localhost:3002" }
    }
  }
}

Deploying (Docker)

cp .env.example .env
# edit .env — at minimum set AGENTGATE_JWT_SECRET to a 32+ char secret
docker compose up --build

The API refuses to boot in production if AGENTGATE_JWT_SECRET still has the dev default.

Data persists in the agent-tools-data named volume (sqlite db).

Postgres path

We use SQLite for v0 (single-writer, fine for low traffic, simple ops). Drizzle's schema can switch to pgTable to back the same routes with Postgres when traffic justifies it — that's deliberately out of scope for v0 to keep the deploy story one command.

Testing

npm -w @agent-tools/api test

15 tests covering: developer signup + duplicate handling, anon/authed publish, ownership-bound delete & version, JSON Schema meta validation, runtime input validation, SSRF (IPv4 + IPv6 literals), AgentGate JWT (missing/bogus/expired/wrong-scope/valid), and MCP initialize/list/unknown-method.

License

MIT

Source & license

This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.