TKeeper – policy-governed, signed intents for autonomous systems
Intent-bound cryptographic proofs for AI agents when existing guardrails lack verifiable authorization.
auth for agents–from the creator of `dotenv` and `dotenvx`
Agent auth via key-signing beats API keys and OAuth for autonomous systems.
AI agent developers, API providers, backend engineers building agent-friendly infrastructure
OAuth 2.0 Device Flow · mTLS · JWT signing
I searched for solutions but wasn't happy with any so I created Vestauth.
Here's how it works:
It manages both the agent and the provider side. The agent with one command can set up a cryptographic identity avoiding human designed handshake mechanisms like OAuth. And on the provider side there is no management of API keys, no username and passwords, no users table even. Authentication works with a single line of code verifying this cryptographically.
Intent-bound cryptographic proofs for AI agents when existing guardrails lack verifiable authorization.
Auth0 for AI agents, but the premise assumes a problem that doesn't yet exist.
Payload signing for agents exists—MCP and schema validation already gate execution. Missing: evidence this prevents real exploits.
Ed25519 signed receipts solve AI agent accountability across org boundaries.
Schnorr signatures and ZK proofs for AI agent identity before this becomes a critical problem.
TLS for MCP agents with ECDSA passports and L0-L4 trust levels, zero dependencies.