The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes instructions and scripts that execute shell commands. Specifically, the agent is instructed in SKILL.md to use the find command to locate the skill's installation directory. Furthermore, both eval/run_eval.py and assets/test_collect_evidence.py use subprocess.run to execute the skill's internal evidence collection logic during testing and evaluation.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection because it reads and processes untrusted data from a target repository to generate reports and suggested fixes.
Ingestion points: The assets/collect_evidence.py script walks the target repository and reads the contents of configuration files, build scripts, and documentation into the agent's context.
Boundary markers: Absent. The skill instructions do not specify the use of delimiters or provide warnings for the agent to ignore instructions embedded in the audited files.
Capability inventory: The skill can execute local shell commands to run the collector and possesses file-writing capabilities to perform repository scaffolding in Step 5.
Sanitization: Absent. There is no mention of sanitizing or validating the extracted repository content before the agent uses it to generate audit findings or code modifications.

agent-ready-rails