The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from multiple sources (file paths, pasted text, and an inbox folder) without explicitly defined sanitization or boundary markers to prevent the AI from following instructions embedded within that data.
Ingestion points: Ingests arbitrary text from input files, pasted text, and the vault's _inbox/ directory.
Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are provided for the ingested content in the instructions.
Capability inventory: The skill utilizes vault_search and vault_reindex, and it has the capability to create, modify, and delete files within the vault.
Sanitization: The instructions do not specify any escaping, validation, or filtering of the external content before processing.
[COMMAND_EXECUTION]: The skill instructs the agent to execute a bash command to check for the presence of a configuration overlay directory using an environment variable.
Evidence: [ -n "$POWERSUIT_OVERLAY" ] && ls -1 "$POWERSUIT_OVERLAY/vault" 2>/dev/null || true in SKILL.md.

atomize