The Agent Skills Directory

[COMMAND_EXECUTION]: The workflow instructs the agent to execute shell commands (e.g., npm run validate or skills-ref validate) if a validator is present in the repository being reviewed. While appropriate for a developer tool, this involves running local scripts based on the presence of specific files in the target project.
[INDIRECT_PROMPT_INJECTION]: The skill's primary purpose is to ingest and analyze external skill definitions (SKILL.md and referenced files), which creates an attack surface for indirect prompt injection.
Ingestion points: The workflow specifically requires reading and summarizing the content of external SKILL.md files and all linked dependencies.
Boundary markers: The instructions do not specify any delimiters or safety markers to help the agent distinguish between the reviewer's instructions and the content being reviewed.
Capability inventory: The skill allows for command execution (validation scripts) and file modification (applying fixes), which could be targets for manipulation by malicious content inside a reviewed skill.
Sanitization: There are no explicit instructions for sanitizing or escaping the content read from external files before the agent processes or summarizes it.

skill-quality-reviewer