Haldir MCP Server

The open-source governance layer for AI agents. Identity, secrets, audit, and policy enforcement - MIT licensed, self-host or use our cloud. It runs locally over stdio via the published package.

People connecting security tools to Claude, Cursor, VS Code, or another MCP client. The project is written in Python.

VERIFIED ACTIVE

LAST COMMIT 2026-04-22 · ★ 4 · #127 OF 182 MAINTAINED SECURITY · VERIFIED 2026-08-25

MIT · Python servers · how we verify → /methodology

01 · Install Haldir

before you install - you'll need

HALDIR_API_KEY

Environment variables documented in the project's README - it lists which are required and which have defaults.

Claude Code

claude mcp add exposureguard-haldir -- uvx haldir

Claude Desktop / Cursor / VS Code - add to config

{
  "mcpServers": {
    "exposureguard-haldir": {
      "command": "uvx",
      "args": [
        "haldir"
      ]
    }
  }
}

Same JSON for Cursor. For VS Code, rename the top-level key from `mcpServers` to `servers`.

Using another client? Same JSON, different key

Claude Desktop · mcpServers

Cursor · mcpServers

VS Code · servers

Windsurf · mcpServers

Zed · context_servers

Cline · mcpServers

Roo Code · mcpServers

Continue · mcpServers

LibreChat · mcpServers

Gemini CLI · mcpServers

Codex CLI · mcp_servers

Full setup guides: every client.

02 · Evidence

Security posture

What to check before giving this server access to your agent - from the registry, GitHub, and our own probes. We don't score safety; we show what's verifiable.

runs as local process (stdio) - runs on your machine with your user's permissions

license MIT - declared in the repository

pypi package haldir - check the name against the project README before installing (PyPI has no namespace ownership)

registry namespace io.github.ExposureGuard is GitHub-verified and matches the repo owner

03 · What Haldir can do

Prose above is summarized from the project's README and registry record - no invented capabilities.

Latest releases

v0.2.2 · 2026-04-18

Big 4 agent-framework integrations complete · Haldir now ships native governance for every major AI agent framework developers use in 2026: · langchain-haldir · crewai-haldir · autogen-haldir - new in this release…

v0.2.1 · 2026-04-18

Vault upgraded to AES-256-GCM · AES-128-CBC + HMAC-SHA256 (Fernet). · AES-256-GCM with additional authenticated data (AAD) binding each ciphertext to its (tenant_id, secret_name). · What this means for buyers…

v0.2.0 · 2026-04-18

Haldir v0.2.0 - now embedded in every major agent framework · Native integrations shipped · langchain-haldir - governance callbacks, tool wrappers, secrets helper · crewai-haldir - CrewAI BaseTool wrapping with scope…

04 · Who maintains Haldir

Haldir is maintained by exposureguard. We track 2 MCP servers from exposureguard - 2 actively maintained, 4 combined GitHub stars, oldest repo from Mar 2026.

05 · Facts

category
security - ranked #127 of 182 actively-maintained security servers as of 2026-08-25.
registry
io.github.ExposureGuard/haldir (active, first published 2026-04-05)
packages
pypi:haldir

06 · Haldir FAQ

What is Haldir?

The open-source governance layer for AI agents. Identity, secrets, audit, and policy enforcement - MIT licensed, self-host or use our cloud. It runs locally over stdio via the published package.

Is Haldir still maintained?

Yes - as of 2026-08-25, its last commit was 2026-04-22. We re-verify nightly.

How do I install Haldir?

Run `uvx haldir`. The README documents one environment variable (HALDIR_API_KEY) to set first. You can also paste the ready-made client config above.

Does Haldir run locally?

Yes - it's a stdio server: it runs on your machine (via uvx) with your user's permissions. Your data stays local unless the server itself calls external APIs.

07 · Alternatives to Haldir