The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill implements an 'Overlay' mechanism that dynamically loads instructions from a path specified by the $POWERSUIT_OVERLAY environment variable. It instructs the agent to read a README.md and subsequent files in that directory, and explicitly mandates that the instructions found in the overlay take precedence over the base skill instructions. This allows for arbitrary behavior modification based on external file content.
[COMMAND_EXECUTION]: The skill requires the agent to execute a shell command [ -n "$POWERSUIT_OVERLAY" ] && ls -1 "$POWERSUIT_OVERLAY/vestige" 2>/dev/null || true to probe the file system for the presence of configuration directories.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data (pasted prose or files specified by path) for style auditing. This creates a vulnerability surface where malicious instructions embedded in the audited text could attempt to influence the agent's behavior.
Ingestion points: File paths provided to the /style-review command and raw text pasted by the user.
Boundary markers: The skill does not define clear delimiters or use 'ignore embedded instructions' warnings when processing input text.
Capability inventory: The skill possesses file reading capabilities and shell command execution (ls) for environment discovery.
Sanitization: No sanitization or filtering logic is present for the input text before it is processed by the agent.

style-review