The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute local Python scripts included in the package to perform its primary analysis functions.
Evidence: SKILL.md workflow steps command the agent to run python scripts/validate_skill.py <skill-path> and python scripts/analyze_skill_body.py <skill-path> --json.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze the content of other agent skill files (SKILL.md), which constitutes an attack surface for instructions embedded in external data.
Ingestion points: The scripts scripts/validate_skill.py and scripts/analyze_skill_body.py read and parse the content of SKILL.md files from directories specified by the user.
Boundary markers: The instructions do not mandate the use of specific boundary delimiters or 'ignore' instructions when the agent processes the analysis output, which could lead to instructions in the analyzed skill influencing the agent's behavior.
Capability inventory: The skill possesses file system read capabilities via its Python scripts and the ability to execute these scripts through the agent.
Sanitization: The skill implements security best practices for parsing by using yaml.safe_load() for YAML frontmatter and restricted regular expression matching for body analysis, rather than executing or evaluating the analyzed content directly.

engineer-skill