The Agent Skills Directory

[SAFE]: The skill facilitates document analysis using standard platform tools like readDocument and parseExcel for various file formats. No patterns for data exfiltration, command execution, or obfuscation were detected.\n- [SAFE]: The skill handles untrusted user-provided content which is a surface for indirect prompt injection. Ingestion points: readDocument and parseExcel tool calls in SKILL.md. Boundary markers: Absent. Capabilities: Document reading and summarization. Sanitization: Absent. The logic is restricted to the skill's stated purpose and contains no instructions to bypass agent safety guidelines.

analyze-document