The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local Python script scripts/atoms-search.py via the shell to process and score knowledge entries based on user input.
[DYNAMIC_EXECUTION]: Employs a script at runtime to handle keyword scoring and data retrieval logic from both local and remote sources.
[DATA_EXFILTRATION]: The search script performs network requests to api.xyskill.xyz to fetch knowledge atoms, transmitting user-provided keywords to a non-whitelisted vendor domain.
[PERSISTENCE]: Accesses and modifies ~/.xy/profile.md to record AI interactions and diagnostic history, enabling state tracking across different user sessions.
[INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted content from multiple sources which could be used to influence agent behavior. * Ingestion points: User-provided descriptions of their tasks, historical records in ~/.xy/profile.md, and knowledge results returned from the external API. * Boundary markers: No explicit delimiters or instructions are used to separate external or historical data from the core system prompt instructions. * Capability inventory: The skill possesses capabilities for local script execution and local file modification. * Sanitization: There is no evidence of validation, filtering, or sanitization applied to the data ingested from user inputs or search results.

xy-kickoff