The Agent Skills Directory

[PROMPT_INJECTION]: The skill identifies project types by reading metadata from codebase files such as theme headers and JSON configurations, which constitutes a potential indirect prompt injection surface.
Ingestion points: The agent is instructed in SKILL.md and references/decision-tree.md to read files such as style.css, theme.json, composer.json, and package.json from the repository.
Boundary markers: No explicit boundaries or instructions are provided to the agent to ignore potentially malicious instructions embedded in these files.
Capability inventory: The agent has permissions to execute bash and node commands, including the execution of build, lint, and test scripts identified in the repository.
Sanitization: The skill does not define sanitization or validation steps for the content read from these external files before using it to influence routing decisions or tool execution.
[COMMAND_EXECUTION]: The skill procedure includes executing build, lint, and test scripts found within the repository's configuration. It also suggests the use of the skills command-line tool via npx to discover additional domain-specific skills.

wordpress-router