| name | paxeer-identity |
| description | Establish and prove the agent's Paxeer network identity — resolve/provision the embedded wallet (network-side custody), bind its address to a durable cortex Identity, report the on-chain footprint, and sign messages as proof. Read + sign only; never moves value. |
| origin | Matrix/Paxeer |
Paxeer Identity
L1 of the Paxeer machine-commerce stack: agent identity. Before an agent
can pay, trade, stake, or schedule on Paxeer, it needs a wallet and a durable
sense of who it is on the network. This skill resolves that identity and
writes it into cortex so every other paxeer-* skill can act under it.
Custody model (why this is safe)
The agent's wallet is an embedded wallet at connect.paxportwallet.com.
Keys live with the network custody service and signing happens server-side —
the bridge never sees key material. Because custody is network-side, spending
limits, allow-lists, and policy are enforced on the wallet at the network
layer, not trusted to agent code. That is precisely why this is the right
identity surface for autonomous agents acting on behalf of users.
Tool mandate (the planner MUST follow)
- ALWAYS resolve the address with a
paxeer-net.wallet_info tool_call. It
provisions the wallet on first use and returns the authoritative address
chainId. NEVER invent or guess an address.
- To establish identity on behalf of an owner, persist a cortex
Identity
memory binding the owner's name/DID to the resolved address
(Wallets:[address]) + chain. This binding is what lets ,
, etc. act under the right identity.
