The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external scientific sources, creating a potential attack surface.\n- Ingestion points: Untrusted data enters the agent context through API responses in SKILL.md (via requests.get().json()) and CSV file loading in SKILL.md and references/dependency_analysis.md (via pd.read_csv()).\n- Boundary markers: Absent. The instructions do not define specific delimiters or warnings to ignore instructions that might be embedded in the scientific data.\n- Capability inventory: The skill utilizes network access (requests.get) to fetch data and file writing (open()) to save downloads to the local system in SKILL.md.\n- Sanitization: Absent. There is no evidence of validation or filtering applied to the content of the API responses or CSV files before they are processed by the agent.

depmap