AGENTS / GITHUB / haldir
githubinferredactive

haldir

provenance:github:ExposureGuard/haldir
WHAT THIS AGENT DOES

haldir is an agent designed to provide a guardian layer for AI agents, focusing on identity, secrets management, and auditing through the Model Context Protocol (MCP). It integrates components like Gate, Vault, and Watch to enhance security. This agent is useful for developers and organizations deploying AI agents who need robust governance and security controls. haldir aims to ensure compliance and provide a secure environment for AI agent operations, potentially leveraging tools like CrewAI and Langchain.

PROBLEM IT SOLVES

haldir solves the problem of managing identity, secrets, and audit trails for AI agents, which can be complex and error-prone when handled manually. It provides a centralized and automated solution for these critical security and governance aspects, reducing the risk of unauthorized access and ensuring compliance.

View Source ↗First seen 2mo agoNot yet hireable

CAPABILITIES & CONSTRAINTS

TECH & STACK
pythonai-agentssecuritygovernancelangchainmcp

PUBLIC HISTORY

First discoveredApr 19, 2026

IDENTITY

inferred

Identity inferred from code signals. No PROVENANCE.yml found.

Is this yours? Claim it →

METADATA

platformgithub
first seenApr 3, 2026
last updatedApr 19, 2026
last crawled21 days ago
version

README BADGE

Add to your README:

![Provenance](https://getprovenance.dev/api/badge?id=provenance:github:ExposureGuard/haldir)