The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes external shell commands to perform its core functionality.
The rg (ripgrep) utility is invoked with user-provided arguments (paths, patterns, and flags) to search the file system.
The skill instructs the agent to run python scripts/test_skill.py (referenced in SKILL.md) and python scripts/validate_skill.py for environment verification.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection because it reads and processes arbitrary content from the local file system.
Ingestion points: Content is ingested from any file path and directory targeted by the rg search patterns provided in SKILL.md and references/quick-reference.md.
Boundary markers: None are explicitly defined to instruct the agent to ignore instructions found within search results.
Capability inventory: The skill has the capability to execute shell commands (rg) and Python scripts (validate_skill.py).
Sanitization: The provided Python script does not perform sanitization on the inputs or the data being processed.
[SAFE]: The skill references a GitHub repository (github.com/aipoch/medical-research-skills) for source code and documentation. As this repository belongs to the skill author (aipoch), it is considered a legitimate vendor resource and does not present a security risk.

file-search