The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to search for identity keys using a bash command that scans both the local project directory and the user's home directory (specifically ~/.claude/*.key).\n- [DATA_EXFILTRATION]: The discovery of sensitive private key files represents a risk of credential exposure, as these materials could be leaked if the agent logs or returns the output of the search command.\n- [PROMPT_INJECTION]: The skill identifies a workflow for deploying and testing external pattern files (packages/patterns/[name]/main.tsx), which creates an entry point for indirect prompt injection.\n
Ingestion points: Content is read from packages/patterns/[name]/main.tsx during deployment and testing.\n
Boundary markers: The skill does not provide delimiters or instructions to ignore embedded prompts within the pattern files.\n
Capability inventory: The agent uses deno task to execute, update, and test patterns, which may involve network and system-level access.\n
Sanitization: There is no evidence of validation or sanitization of the code patterns before they are processed by the CLI.

pattern-deploy