mishrasanjeev

We reviewed 30 of the most popular AI agent projects on GitHub (OpenClaw, AutoGen, CrewAI, LangGraph, MetaGPT, AutoGPT, etc.) across six authorization criteria: scoped permissions, per-agent identity, user consent, revocation, audit trails, and delegation control. Key findings: - 93% rely on unscoped API keys as the only auth mechanism - 0% have per-agent cryptographic identity -...

https://github.com/mishrasanjeev/grantex

AI agents are being deployed to schedule meetings, write code, manage infrastructure, and transact on behalf of users. But there's no standard way to authorize what an agent can do, audit what it did, or revoke access when something goes wrong. Most teams are hacking together API keys and custom RBAC — none of it is portable, auditable, or revocable at the agent level. Grantex is an...

https://github.com/mishrasanjeev/grantex