mukundakatta avatar

Agentguard

mukundakatta/agentguard-mcp

A policy checker for URLs that agents want to fetch, surfaced as three MCP tools: check_url for single decisions, check_urls_batch for bulk analysis, and validate_policy to catch common mistakes like overly broad wildcards or malformed host patterns. The server wraps the @mukundakatta/agentguard library so Claude Desktop, Cursor, or any MCP client can ask "would my firewall allow this?" before letting a tool make a real request. Note that this only returns allow/deny verdicts; it doesn't intercept or block actual network calls. Useful when you're writing egress policies and want the LLM to help audit which URLs would pass, or when prototyping multi-tool workflows where network boundaries matter.

GitHub stars
1
tools
3
transport
STDIO
last updated
Apr 27, 2026

Tools

3 tools

Public tool metadata for what this MCP can expose to an agent.

check_urlSingle URL check: returns { allowed, reason } without making any actual request.

Single URL check: returns { allowed, reason } without making any actual request.

No parameter schema in public metadata yet.

check_urls_batchBatch check with per-URL decisions plus a summary.

Batch check with per-URL decisions plus a summary.

No parameter schema in public metadata yet.

validate_policySanity-check a policy spec for empty allowlists, overly broad * wildcards, and malformed host patterns.

Sanity-check a policy spec for empty allowlists, overly broad * wildcards, and malformed host patterns.

No parameter schema in public metadata yet.

Sponsors

Readme

agentguard-mcp

MCP server for @mukundakatta/agentguard. Lets Claude Desktop, Cursor, Cline, Windsurf, Zed, or any other MCP client check whether a URL is allowed under a network-egress policy before any fetch.

npx -y @mukundakatta/agentguard-mcp

Three tools:

  • check_url — single URL check: returns { allowed, reason } without making any actual request.
  • check_urls_batch — batch check with per-URL decisions plus a summary.
  • validate_policy — sanity-check a policy spec for empty allowlists, overly broad * wildcards, and malformed host patterns.

Add to your client

Claude Desktop

Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "agentguard": {
      "command": "npx",
      "args": ["-y", "@mukundakatta/agentguard-mcp"]
    }
  }
}

Same shape for Cursor (~/.cursor/mcp.json), Cline, Windsurf, Zed.

Tool examples

check_url:


View the full README on GitHub

Details

Repository
github.com/mukundakatta/agentguard-mcp
Website
github.com/MukundaKatta/agentguard-mcp
Categories
AI & LLM Tools, Security & Pentesting
Package
@mukundakatta/agentguard-mcp
Transport
STDIO
Registry
MCP registry, active

More from mukundakatta

Related AI & LLM Tools MCP servers

View all
1mukundakatta avatarAgentvetmukundakatta/agentvet-mcpValidate tool-call args before execution. Returns LLM-friendly retry hints.1stars
2new1direction avatarKorg Ledgernew1direction/korgexVerifiable cognition: prove a session ledger is tamper-evident; audit & import any agent's logs.1stars
3agentra-commerce avatarXapagentra-commerce/xap-sdkAgent settlement protocol — discover, negotiate, settle, and verify with XAP.0stars
4pquattro avatarMemoraEUpquattro/memoraeu-memoraeu-mcpZero-knowledge persistent memory for Claude — encrypted, semantic search, EU hosted.0stars
5u2n4 avatarVEO 3.1u2n4/gemini-media-mcp4K video generation via Google VEO 3.1: text-to-video, image-to-video, video extension0stars
6ivanantigravity-lgtm avatarModel Council Mcp Serverivanantigravity-lgtm/model-council-mcp-serverThree-model MCP server returning Grok, Gemini, and DeepSeek outputs for Claude.1stars