The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process threat intelligence data (such as indicators, event info, and tags) from external MISP servers, which act as untrusted ingestion points.
Ingestion points: Data enters the context via MISPClient.search_attributes, MISPClient.search_events, and MISPClient.get_event in scripts/agent.py.
Boundary markers: The instructions do not define boundary markers to separate trusted instructions from external data content.
Capability inventory: The skill uses urllib.request for API communication and performs data correlation; no high-risk capabilities like file system modification or shell subprocess execution were found.
Sanitization: There is no evidence of sanitization or schema validation for the data retrieved from external APIs.
[EXTERNAL_DOWNLOADS]: The scripts/agent.py script communicates with a configurable external MISP API. The implementation in MISPClient defaults to disabling SSL verification (ssl.CERT_NONE), creating a risk for interception of security data in transit.
[COMMAND_EXECUTION]: The skill includes a standalone script scripts/agent.py that executes search and correlation logic based on command-line arguments.

correlating-threat-campaigns