@0n1x/rhinogent
# @0n1x/rhinogent
**The identity wallet for AI agents — the front door to the 0n1x trust ecosystem.**
Rhinogent gives an agent a self-custody identity (key generated client-side, never custodied) and an
**MCP server** that plugs it into [0n1x](https://onyx-actions.onrender.com) — the trust layer where
agents verify counterparties, sign facts, carry portable reputation, and transact safely with external
agents. Wallets are commoditized; **Rhinogent is the identity + reputation layer that makes a wallet
worth trusting.**
> Everyone secures your agent's keys. Rhinogent also tells it *who it's paying.*
## The picture
```
your agent (Claude / Cursor / Codex / your code)
│ holds its own key (Rhinogent wallet, client-side)
▼
@0n1x/rhinogent (this MCP server / SDK) ←—— the front door
│ one connection = your passport
▼
0n1x ecosystem: verify · sign-facts · identity · reputation · mandates · data + financial exchange
│
▼
the agentic web — transact + interact with EXTERNAL agents, trust verified
```
Hold a Rhinogent identity → you're a citizen of 0n1x → you get the perks: the security, data, and
financial tools, plus the trust rails to deal with agents you don't control.
## Install (add to any MCP client in one line)
```jsonc
// claude_desktop_config.json (or Cursor / Codex mcp config)
{
"mcpServers": {
"rhinogent": { "command": "npx", "args": ["-y", "@0n1x/rhinogent"] }
}
}
```
## Tools
| Tool | What it does |
|---|---|
| `rhinogent_mint_identity` | Onboard a fresh self-custody agent → signed identity card (callsign, did:pkh, 0n1x-Verified) |
| `rhinogent_identity` | Resolve an address → its public signed identity |
| `rhinogent_verify_counterparty` | **Verify before you pay** — signed verdict on a merchant/domain, grounded in facts |
| `rhinogent_sign_fact` | **Sign a fact** (not a judgment) — Ed25519 attestation of verifiable reality |
| `rhinogent_get_reputation` | The intelligent, un-gameable reputation — proper-scoring + stake + behavior |
| `rhinogent_mandate` | Confirm a spend/action is within the agent's granted mandate |
Keys never touch this server — it only calls the public, signed 0n1x trust endpoints.
Point at a different backend with `ONYX_BASE`.
## Why it matters
The industry calls agent **identity + reputation** the *unsolved frontier* — every wallet provider
ships keys, none ship *trust*. Rhinogent does: signed facts, portable reputation, verify-before-you-pay.
That's the one thing that turns an empty wallet into an agent other agents (and merchants) will deal with.
— built on 0n1x · self-custody · neutral · signed + replayable
TDQS
Scored across 6 tools
Each tool targets a distinct function: identity creation vs resolution, reputation reading, mandate checking, fact signing, and counterparty verification. No overlap in purpose or ambiguity.
All tools share the 'rhinogent_' prefix and use snake_case, but naming pattern is mixed: some use action verbs (get, mint, sign, verify) while others are nouns (identity, mandate). Still clear and predictable.
With 6 tools, the server covers core agent identity and trust operations without being over- or under-scoped. Each tool serves a clear, necessary purpose.
Covers create/read for identity, read for reputation, and verification actions. Minor gaps: no update/delete for identity, no mandate creation or outcome submission tools, but these may be handled externally.