deep-document-processor
Pass
Audited by Gen Agent Trust Hub on Apr 2, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill defines a text-processing ingestion surface for indirect prompt injection but is safe as it lacks dangerous capabilities. 1. Ingestion points: External documents, codebases, or research papers provided as input (SKILL.md). 2. Boundary markers: Uses structural delimiters like '=== CONTEXT: [Document Title] ===' to frame outputs. 3. Capability inventory: None; the skill contains no executable scripts, tool calls, or network functions. 4. Sanitization: None present.
- [NO_CODE]: No code or scripts were found in the skill. All logic is provided as natural language instructions.
Audit Metadata