agentic-store-mcp 0.4.0

Open-source MCP tools for AI agents — lint, security, dependency audit, web scraping and more

agentic-store-mcp

Give your AI agents superpowers — free, open-source MCP tools, self-hosted, no account required.

Works with Claude Desktop, Cursor, Windsurf, VS Code, and any MCP-compatible client.

---

What's Included

Module	Tool	Description	Requires
AgenticCode	python_lint_checker	Static analysis for Python — bugs, style, complexity	Nothing
AgenticCode	repo_scanner	Scan for leaked secrets, PII, and .gitignore gaps	Nothing
AgenticCode	dependency_audit	Check for outdated and vulnerable dependencies	Nothing
AgenticData	agentic_web_crawl	Fetch any URL — extract page text, SEO metadata, headings, links, images	Nothing
AgenticData	agentic_web_search	Search the web via self-hosted SearXNG; returns ranked results with title, URL, snippet	SearXNG running

---

Setup

Choose your preferred setup method. Both expose the same tools over the same MCP stdio protocol.

• Python — requires uv (recommended for developers)
• Docker — requires Docker, no Python needed on the host

---

Python Setup

1. Install uv

macOS / Linux

curl -LsSf https://astral.sh/uv/install.sh | sh

Windows (PowerShell)

powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

Windows (winget)

winget install --id=astral-sh.uv -e

Verify:

uv --version

2. Run the server

Option A — Run directly via uvx (no install, always latest)

uvx --refresh agentic-store-mcp

--refresh ensures uvx always pulls the latest published version on each run. Omit it if you want to reuse the cached version.

Option B — Install globally

pip install agentic-store-mcp
agentic-store-mcp

Option C — Clone and run from source

git clone https://github.com/agenticstore/agentic-store-mcp
cd agentic-store-mcp
uv sync
uv run server.py

3. Enable web search (optional)

agentic_web_search requires a running SearXNG instance — a free, self-hosted metasearch engine.

# Start SearXNG (Docker required)
docker compose up -d

# Run the MCP server with the SearXNG URL
SEARXNG_URL=http://localhost:8080 uvx --refresh agentic-store-mcp

4. Available flags

uvx --refresh agentic-store-mcp                              # all tools
uvx --refresh agentic-store-mcp --modules code               # AgenticCode only
uvx --refresh agentic-store-mcp --modules data               # AgenticData only
uvx --refresh agentic-store-mcp --modules code,data          # multiple modules
uvx --refresh agentic-store-mcp --tools agentic_web_search   # single tool
uvx --refresh agentic-store-mcp --list                       # list tools and exit

---

Docker Setup

1. Pull the image (recommended)

docker pull agenticstore/agentic-store-mcp:latest

2. Verify

docker run --rm agenticstore/agentic-store-mcp:latest --list

3. Enable web search (optional)

Requires a running SearXNG instance. Clone the repo just for the docker-compose.yml, then start SearXNG:

git clone https://github.com/agenticstore/agentic-store-mcp
cd agentic-store-mcp
docker compose up -d

Run the MCP server connected to SearXNG:

docker run -i --rm \
  --network agentic-store-mcp_default \
  -e SEARXNG_URL=http://searxng:8080 \
  agenticstore/agentic-store-mcp:latest

Update to latest (optional)

docker pull agenticstore/agentic-store-mcp:latest

Build from source (optional — for contributors or custom builds)

git clone https://github.com/agenticstore/agentic-store-mcp
cd agentic-store-mcp
docker build -t agentic-store-mcp .
docker run --rm agentic-store-mcp --list

Replace agenticstore/agentic-store-mcp:latest with agentic-store-mcp in any config below if using a local build.

---

Connect to Your AI Client

Add AgenticStore to your MCP client's config file, then restart the client.

Each section shows Python and Docker variants. Docker configs use the remote image — if you built locally, replace agenticstore/agentic-store-mcp:latest with agentic-store-mcp.

---

Claude Desktop

Config location

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

Python — all tools

{
  "mcpServers": {
    "agentic-store-mcp": {
      "command": "uvx",
      "args": ["--refresh", "agentic-store-mcp"]
    }
  }
}

Python — with web search (optional — docker compose up -d required)

{
  "mcpServers": {
    "agentic-store-mcp": {
      "command": "uvx",
      "args": ["--refresh", "agentic-store-mcp"],
      "env": {
        "SEARXNG_URL": "http://localhost:8080"
      }
    }
  }
}

Docker — web crawl only

{
  "mcpServers": {
    "agentic-store-mcp": {
      "command": "docker",
      "args": ["run", "-i", "--rm", "agenticstore/agentic-store-mcp:latest"]
    }
  }
}

Docker — with web search (optional — docker compose up -d required)

{
  "mcpServers": {
    "agentic-store-mcp": {
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "--network", "agentic-store-mcp_default",
        "-e", "SEARXNG_URL=http://searxng:8080",
        "agenticstore/agentic-store-mcp:latest"
      ]
    }
  }
}

---

Cursor

Config location: ~/.cursor/mcp.json

Python — all tools

{
  "mcpServers": {
    "agentic-store-mcp": {
      "command": "uvx",
      "args": ["--refresh", "agentic-store-mcp"]
    }
  }
}

Python — with web search (optional — docker compose up -d required)

{
  "mcpServers": {
    "agentic-store-mcp": {
      "command": "uvx",
      "args": ["--refresh", "agentic-store-mcp"],
      "env": { "SEARXNG_URL": "http://localhost:8080" }
    }
  }
}

Docker — with web search (optional — docker compose up -d required)

{
  "mcpServers": {
    "agentic-store-mcp": {
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "--network", "agentic-store-mcp_default",
        "-e", "SEARXNG_URL=http://searxng:8080",
        "agenticstore/agentic-store-mcp:latest"
      ]
    }
  }
}

---

Windsurf

Config location: ~/.codeium/windsurf/mcp_config.json

Same pattern as Cursor above.

---

VS Code (MCP extension)

Same pattern as Cursor above.

---

Programmatic Usage (Python)

from agentic_store_mcp import start_server

start_server()                              # all tools
start_server(modules=["code"])              # AgenticCode only
start_server(modules=["code", "data"])      # multiple modules
start_server(tools=["agentic_web_search"])  # single tool

---

Tool Reference

agentic_web_crawl

Fetches a URL and returns structured page content and SEO signals: title, meta description, canonical URL, robots directives, Open Graph tags, Twitter Card tags, heading structure (h1–h6), word count, and internal/external link breakdown. Handles redirects and encoding detection automatically.

No external dependencies or setup required.

agentic_web_search

Searches the web via a self-hosted SearXNG instance. Returns ranked results with title, URL, snippet, and source engine. Supports categories (general, news, images, science, it, etc.) and language filtering.

Requires SearXNG: docker compose up -d

python_lint_checker

Static analysis for Python files. Checks imports, bugs, style, and complexity — zero external dependencies.

repo_scanner

Scans a directory for leaked secrets (AWS keys, API tokens, private keys), PII (emails, SSNs), and missing .gitignore entries.

dependency_audit

Audits requirements.txt, pyproject.toml, package.json, go.mod, pom.xml, and more. Checks for outdated versions and known CVEs via OSV.

---

Contributing

• Tools must be pure Python — no managed services, no AgenticStore infrastructure
• Each tool lives in agentic_store_mcp/modules/<category>/<tool_name>/ with handler.py, schema.json, README.md, and test_handler.py
• Run tests: uv run pytest

---

License

MIT — free to use, fork, and modify.