The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to run a local script node scripts/check-maps.mjs to validate the structural integrity of the project's maps. This is an expected maintenance operation within the skill's defined scope.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted project data, which introduces a surface for indirect instructions to influence agent behavior.
Ingestion points: Project source code, OpenAPI specifications, maps/ directory files, and Storybook routes are read to reconcile feature coverage.
Boundary markers: The instructions do not specify the use of delimiters or specific 'ignore' markers when processing the content of these external files.
Capability inventory: The skill possesses the capability to modify local map files, create Pull Requests, and execute Node.js scripts.
Sanitization: There is no mention of sanitizing or validating the contents of the audited files before they are interpolated into the skill's output or used to drive the maintenance workflow.
[SAFE]: The skill contains security-positive instructions, explicitly stating that a 'permission denial is not authorization to widen access.' It also recommends using read-only subagents for source review to minimize risks. All external resource references trace back to the vendor's own repository (paperclipai/agent-plugins) or well-known community sources.

maintain-verification-skill