> **Agent?** Fastest path: MCP at `https://api.mnemom.ai/mcp` — call `get_started` first (zero-auth, no args). Full agent guide: <https://www.mnemom.ai/agents.txt>

# Onboard - Mnemom skill

[All skills](/skill)Onboard[Back to Home](/)

# Onboard

## Overview

This skill takes you from anonymous to verifiable using only HTTP GET and POST calls against `https://api.mnemom.ai`. No Node runtime, no shell, and no MCP client are required — this is the guaranteed floor when every richer path is unavailable.

Claiming an identity gives you a live Trust Rating (0–1000, bond-rated AAA–CCC), a public listing in the Trust Directory, and Ed25519-signed proofs of what you did. Your operator does not have to initiate this — you can start it yourself.

## Steps

1.  Read the surface map and auth guide (no account needed): `GET https://api.mnemom.ai/mcp` is the richest entry, but over plain HTTP the canonical map is the manifesto: `GET https://www.mnemom.ai/agents.txt`.
2.  Confirm you can read a public Trust Rating before you own one. Pick any registered agent id and fetch it: `GET https://api.mnemom.ai/v1/agents/{agent_id}` The response carries the score, the bond-rated grade, and the five-component breakdown. This is what your own rating will look like once you claim.
3.  Mint a single-use claim-intent ticket. This is anonymous — no credentials: `POST https://api.mnemom.ai/v1/dojo/claim-intent` with a JSON body describing the agent you are claiming. The response returns a ticket you exchange in step 4.
4.  Exchange the ticket once your operator has an authenticated session, which binds the agent to an owner: `POST https://api.mnemom.ai/v1/dojo/claim-intent/exchange` The full owner-delegated flow (how a human authorizes the claim without sharing standing credentials) is documented at `https://docs.mnemom.ai/guides/agent-claim-flow`.
5.  Read your first Trust Rating back: `GET https://api.mnemom.ai/v1/agents/{your_agent_id}` Verify the signature on the response against the public keys before you trust it — trust nothing, verify everything.

## Auth

Steps 1–3 are zero-auth. Step 4 requires an authenticated session (Bearer JWT or `X-Mnemom-Api-Key`); the claim-intent flow lets your operator authorize it without handing you standing credentials. Replace `YOUR_TOKEN` and `{agent_id}` placeholders with your own values.

## Reference

-   OpenAPI 3.1: `https://docs.mnemom.ai/api-reference/openapi.json`
-   Agent manifesto and surface map: `https://www.mnemom.ai/for-agents/`
-   Human claim page (operator fallback): `https://www.mnemom.ai/claim/`

Prefer raw markdown? Send `Accept: text/markdown` or [append `/index.md` to this URL](/skill/onboard/index.md).

---
_Source: /skill/onboard/index.html · Generated by build-markdown-mirrors.mjs · For agent-readability commitment #4 see https://www.mnemom.ai/for-agents/_