The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute multiple Python scripts (convert_skills.py, remove_duplicates.py, final_cleanup.py) and provides Python code snippets for file manipulation that the agent is expected to run.
[REMOTE_CODE_EXECUTION]: The skill references and instructs the agent to run scripts located in a scripts/ directory that were not provided in the analyzed content, preventing a full security audit of the logic being executed. Referencing and executing unprovided local code constitutes a security risk.
[PROMPT_INJECTION]: The skill processes SKILL.md files which are untrusted external data. This creates a surface for indirect prompt injection where malicious instructions inside the target files could attempt to influence the agent's actions during the standardization process.
Ingestion points: The skill reads content from SKILL.md files in the repository.
Boundary markers: No delimiters or "ignore instructions" warnings are utilized when processing the untrusted file content.
Capability inventory: The agent is granted capabilities to execute Python scripts, use Bash, and perform file write/edit operations.
Sanitization: No evidence of sanitization, validation, or escaping of the file content is present before it is processed.

skill-standardization