The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using uv run to run local Python scripts for vault indexing and monitoring tasks.
[DYNAMIC_EXECUTION]: The skill implements an overlay system that dynamically loads and prioritizes instructions from a directory path specified by the $POWERSUIT_OVERLAY environment variable, allowing the skill's behavior to be modified by external local files.
[INDIRECT_PROMPT_INJECTION]: The skill ingests user input via an interview process and uses this data to generate vault files, which represents an ingestion surface for potentially untrusted data. (1) Ingestion points: Phase 2 interview responses regarding work domains and privacy boundaries. (2) Boundary markers: Absent. (3) Capability inventory: Writing files to the local filesystem and executing shell commands via uv run. (4) Sanitization: Absent; user answers are written directly into configuration files like CLAUDE.md.

seed