The Agent Skills Directory

[SAFE] (SAFE): The skill is entirely composed of documentation and question templates. There are no scripts, binaries, or commands that could be executed maliciously.
[INDIRECT_PROMPT_INJECTION] (LOW): The skill provides instructions for the agent to ingest and analyze external data (e.g., reading business documents, user manuals, and support tickets). This creates a theoretical surface for indirect prompt injection if those documents contain malicious instructions. However, the skill's restricted toolset (Read, Grep, Glob, AskUserQuestion, Write) prevents significant harm as it lacks network or code execution capabilities.
Ingestion points: Read, Grep, and Glob tools used for document analysis of external business files.
Boundary markers: None explicitly defined for separating ingested document content from the agent's instructions.
Capability inventory: Limited to file read/write and user interaction; no shell access or network requests.
Sanitization: No specific sanitization or validation logic is present for the ingested text.

requirements-elicitation