What is 1Claw
1Claw is an HSM-backed, zero-trust vault that lets humans store secrets and lets AI agents fetch only what they're allowed to at runtime.
1Claw solves a specific problem: your API keys, tokens, and private keys end up pasted into .env files, prompts, and CI configs where anything can read them. 1Claw stores them in HSM-backed vaults instead, and hands them out only to principals a human explicitly allowed.
- HSM-backed: each vault's key encryption key lives in a hardware security module
- Zero-trust: an agent gets zero secrets until a human writes a policy
- Two principals: humans own secrets, agents fetch them at runtime
- Same API for both at https://api.1claw.co
Humans authenticate with email/password, Google, passkey, or a personal API key (1ck_). Agents authenticate with an agent API key (ocv_) that they trade for a short-lived JWT.
- 1
Sign up for a free account at the dashboard. The free tier gives you 1,000 requests/month, 3 vaults, and 2 agents.
texthttps://1claw.co - 2
Get a personal API key from Settings → API Keys. It starts with 1ck_. Export it so the next commands can use it. Replace 1ck_your_key_here with the key you copied from the dashboard at Settings → API Keys.
bashexport ONECLAW_API_KEY="1ck_your_key_here" - 3
Verify the key works by exchanging it for an access token against the real API. You'll get back an access_token you can use as a Bearer token.
bashcurl -s -X POST https://api.1claw.co/v1/auth/api-key-token \ -H "Content-Type: application/json" \ -d "{\"api_key\":\"$ONECLAW_API_KEY\"}"
You now have a verified 1Claw account and a working API key. That's the credential every other lesson builds on.
Check your understanding
3 questionsWhat backs the encryption key for each 1Claw vault?
By default, how much access does a new agent have?
Which prefix identifies a human's personal API key?