The Agent Skills Directory

[SAFE]: The skill implements its core functionality using standard Python processing. Analysis of the source code in scripts/cti_utils.py confirms that it performs passive data processing (extraction, normalization, and reporting) without engaging in dangerous operations like arbitrary code execution, unauthorized network access, or persistence mechanisms.\n- [INDIRECT_PROMPT_INJECTION]: \n
Ingestion points: The skill ingests untrusted text data through IOCExtractor.extract_from_text and IOCCollection.add_ioc, typically sourced from external threat reports or logs.\n
Boundary markers: None are explicitly defined in the utility scripts; the skill assumes the calling agent manages data boundaries.\n
Capability inventory: The ATTACKMapper.export_navigator method in scripts/cti_utils.py provides a capability to write JSON data to a user-specified file path.\n
Sanitization: The IOCExtractor.validate_ioc method provides validation of indicator formats using predefined regular expression patterns, reducing the risk of processing malformed or malicious payloads.\n- [EXTERNAL_DOWNLOADS]: The scripts/requirements.txt file mentions requests as an optional dependency but does not include any automated installation commands or active remote code fetches. References to MITRE ATT&CK and STIX specifications in SKILL.md point to well-known, trusted industry resources.

threat-intelligence