The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill performs network requests to ledger.drx4.xyz and ledger-test.drx4.xyz to list, create, and manage trade offers. These domains are not recognized as trusted organizations or well-known services.
[DATA_EXFILTRATION]: The skill retrieves the Bitcoin private key (btcPrivateKey) from the wallet-manager to sign messages using the BIP-137 standard. While the key itself is not transmitted, the resulting signatures and the user's Bitcoin address are sent to the external ledger for every write operation, representing a potential exposure of cryptographic proof of identity to an unverified third party.
[INDIRECT_PROMPT_INJECTION]: The skill fetches trade history and agent lists from the ledger API and prints the raw JSON data to the agent's context. This data, particularly the metadata field which is populated by external users, is not sanitized, creating a surface for indirect prompt injection.
Ingestion points: ledgerGet calls in ordinals-p2p/ordinals-p2p.ts that fetch trade data from the /api/trades and /api/agents endpoints.
Boundary markers: There are no delimiters or instructions provided to the agent to disregard instructions found within the fetched data.
Capability inventory: The skill has access to the user's wallet private keys and the ability to perform network POST operations to the ledger.
Sanitization: The skill uses printJson to output external data directly into the agent's interaction loop without any filtering, escaping, or validation of user-controlled fields.

ordinals-p2p