The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to parse and interpret forensic data from external, potentially malicious sources (LNK files and Jump List artifacts). An attacker who controls these files could embed deceptive metadata (e.g., in command-line arguments or file descriptions) designed to mislead an AI agent or a forensic analyst.
Ingestion points: The scripts scripts/agent.py and scripts/process.py ingest binary data directly from user-provided file paths or directories containing .lnk and .automaticDestinations-ms files.
Boundary markers: No specific boundary markers or instructions are provided to the agent to treat extracted metadata as untrusted or to ignore potential instructions embedded within those fields.
Capability inventory: The skill is capable of reading files, writing JSON reports to the local file system, and displaying extracted information to the terminal.
Sanitization: While the Python scripts use the struct module for structured binary parsing, they do not sanitize text strings (e.g., target paths, arguments) extracted from the link blocks before displaying them.
[EXTERNAL_DOWNLOADS]: The documentation in SKILL.md and references/api-reference.md directs users to install third-party Python packages (LnkParse3, pylnk3) and download external binary tools (Eric Zimmerman's LECmd and JLECmd). While these are standard forensic tools, they are retrieved from external repositories (GitHub and PyPI).

analyzing-lnk-file-and-jump-list-artifacts