The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes MDX files which could potentially contain malicious instructions or hidden payloads designed to influence the behavior of the reviewing sub-agents.
Ingestion points: Documentation files located within the src/content/ directory.
Boundary markers: The skill does not implement specific delimiters or instructions to treat the ingested file content as untrusted data.
Capability inventory: The skill's primary actions involve invoking other agents (docs-reviewer, etc.) to compile a checklist. It lacks high-impact capabilities such as file writes, network exfiltration, or arbitrary code execution.
Sanitization: No evidence of input validation or sanitization is present for the files being analyzed.
[COMMAND_EXECUTION]: The skill utilizes the git status command to programmatically identify modified files in the repository when no specific files are provided by the user. This is a standard and safe use of version control tools for development workflows.
[NO_CODE]: The skill is composed entirely of markdown instructions and YAML metadata, without any included scripts, binaries, or external code dependencies.

review-docs