The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external PRDs and codebase content which may contain instructions that could subvert the architecture review process or influence the agent to perform unauthorized file modifications.
Ingestion points: The skill reads PRD files provided via arguments and scans the local codebase using Glob/Grep.
Boundary markers: There are no instructions to the agent to treat external data with specific delimiters or to ignore potential instructions embedded within the data.
Capability inventory: The skill can write to the filesystem, execute commands via npx, and delegate tasks to subagents.
Sanitization: No sanitization or validation of the input data is performed before processing.
[COMMAND_EXECUTION]: The skill uses npx nx show projects to generate a project map and analyze dependencies, which is a standard procedure for Nx-based monorepos.
[EXTERNAL_DOWNLOADS]: The use of npx may trigger the retrieval of the nx package from the npm registry, representing a download from a well-known service.

architect-refine-critique