The Agent Skills Directory

[DATA_EXFILTRATION]: The skill manages sensitive credentials such as WALLET_PRIVATE_KEY, OPENSERV_API_KEY, and OPENSERV_USER_API_KEY. It documents standard security practices like storing these in .env files and notes that the provision() function automatically handles credential binding and state persistence to .openserv.json.
[COMMAND_EXECUTION]: The skill provides instructions for using the npx @openserv-labs/client deploy command to upload agent code to the OpenServ platform. It also includes examples using npx tsx for running local management and cleanup scripts.
[PROMPT_INJECTION]: As a framework for building agents, the skill naturally creates surfaces for indirect prompt injection by ingesting untrusted data through webhook and x402 triggers.
Ingestion points: External data enters through trigger input schemas defined in SKILL.md and examples/agent.ts.
Boundary markers: The provided examples do not explicitly demonstrate the use of delimiters or 'ignore' instructions for external inputs.
Capability inventory: The agent examples utilize capabilities (e.g., processRequest) which are executed when triggers are fired.
Sanitization: The examples do not specify internal validation or sanitization of input strings before they are processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill relies on official packages from the npm registry, specifically @openserv-labs/client and @openserv-labs/sdk.

openserv-client