The Agent Skills Directory

[SAFE]: The skill operates entirely on local project files and user input to refine documentation. It does not perform any network operations or external data transmissions.
[SAFE]: No obfuscation, persistence mechanisms, or privilege escalation attempts were detected. The logic is transparent and focused on the stated task of specification clarification.
[INDIRECT_PROMPT_INJECTION]: The skill processes external content from a feature specification file and user arguments, creating a surface for potential indirect prompt injection.
Ingestion points: Reads the FEATURE_SPEC file and accepts $ARGUMENTS from the user.
Boundary markers: Instructions use markdown formatting to separate process instructions from the data being analyzed.
Capability inventory: The skill reads from and writes to the local FEATURE_SPEC file; no dangerous system-level capabilities are invoked.
Sanitization: The instructions include guidance for the agent to correctly escape single quotes in user arguments, which helps prevent command injection if the data is used in shell contexts.

clarify