The Agent Skills Directory

[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it is designed to ingest and analyze untrusted data (source code and pull request content) without explicit boundary markers or instructions to ignore embedded commands.
Ingestion points: Source code files (e.g., pages/index.tsx), configuration files (next.config.mjs), and pull request metadata.
Boundary markers: None present in the instructions.
Capability inventory: File system read access, code analysis, and integration with version control systems.
Sanitization: No sanitization of the reviewed content is performed before processing by the agent.
[COMMAND_EXECUTION]: The instructions recommend running npm run lint for local quality assurance, which is a standard development practice for identifying code quality issues.
[SAFE]: The skill encourages secure development practices by instructing the agent to audit environment variable exposure, prevent secret commits, and verify the use of HTML sanitization libraries.
[SAFE]: References to external configuration and vendor-specific integrations (Contentstack) are consistent with the skill's stated purpose and follow legitimate vendor resource naming patterns.

code-review