The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill downloads cybersecurity threat data from the official MITRE Corporation GitHub repository. This is a neutral activity required for its core functionality of mapping threat actor TTPs.
[COMMAND_EXECUTION]: Instructions provide standard commands to install legitimate industry-standard libraries such as stix2 and mitreattack-python.
[INDIRECT_PROMPT_INJECTION]: The skill ingests external JSON data, which is a potential surface for malicious input. 1. Ingestion points: JSON data is loaded from the MITRE GitHub repo in scripts/agent.py. 2. Boundary markers: The skill does not implement specific boundary markers for the data it processes. 3. Capability inventory: The agent has the ability to read from the network and write JSON files locally. 4. Sanitization: The skill relies on standard JSON parsing and does not apply custom sanitization logic to the threat actor data.

analyzing-apt-group-with-mitre-navigator