your agents
are you
not me
every commit, every API call, every push your AI agent makes carries your identity. no separation. no scope. no revocation. when it gets compromised, the attacker is you.
we're building the open-source identity layer so agents can be themselves — short-lived, cryptographically distinct from the humans who deploy them.
agents aren't users — they're machines. they need machine identity, not hacked human identity.
signet
Self-sovereign identity built for humans and machines. Proof-of-possession certs replace steal-and-use bearer tokens for commit signing, GitHub Actions, HTTP auth, and more.
github →notme + apas
notme.bot applies self-sovereign identity to AI agents. APAS is the spec for representing what the agent did.
summary → draft spec →| today | with notme | |
|---|---|---|
| identity | your GitHub PAT | agent's own Ed25519 cert (ML-DSA-44 ready) |
| scope | all your repos | orchestrator-enforced, per-task |
| lifetime | until you rotate | hours, not months |
| revocation | rotate your token | near-real-time edge revoke |
| audit trail | nothing | signed commits + chain hash (DSSE signing coming) |
| on compromise | attacker is you | not me |
help build this
open-source and early. we need people who think in Ed25519, OIDC claims, and attestation chains. or people who've had an agent commit as them and felt sick about it.