The Agent Skills Directory

[SAFE]: The skill facilitates interaction with the OpenServ Launch API (https://instant-launch.openserv.ai) to deploy ERC-20 tokens, create liquidity pools, and list existing tokens on the Base Mainnet. The behavior matches the described purpose.
[EXTERNAL_DOWNLOADS]: The skill references standard Node.js packages such as axios, zod, and dotenv, as well as vendor-specific libraries @openserv-labs/sdk and @openserv-labs/client. These are well-known or vendor-owned resources and do not pose a security risk in this context.
[COMMAND_EXECUTION]: Documentation and examples include standard commands for dependency installation (npm install) and platform-specific skill management (npx skills). No malicious or unauthorized command execution patterns were identified.
[INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided data (token name, symbol, description) through its capabilities. However, these ingestion points are protected by Zod schema validation, and the data is limited to the scope of a structured API request to a vendor endpoint. The capability inventory is restricted to network requests (axios), and there is no unsafe interpolation of untrusted data into critical system commands.

openserv-launch