The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands such as find, mkdir, cp, and ls. These are employed for discovering related skills within the environment, managing local documentation files, and checking for the presence of optional private configuration overlays defined by environment variables.
[DYNAMIC_EXECUTION]: A Python script is executed via a shell heredoc to parse file signals from a local template and perform file system globbing to identify which technical modules are already implemented in the repository.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from the audited repository and potentially from private configuration overlays to generate its recommendations, which constitutes an attack surface for indirect instructions.
Ingestion points: Repository code and documentation files (located via glob patterns) and private overlay files (referenced via $POWERSUIT_OVERLAY).
Boundary markers: None identified in the prompt templates.
Capability inventory: Shell command execution (via find, ls, etc.) and writing to the local file docs/fit.md.
Sanitization: No specific filtering or escaping is applied to the ingested content before it is processed for analysis.

estate-assessment