legal-qa-extractor
Pass
Audited by Gen Agent Trust Hub on Apr 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation techniques, or unauthorized data access behaviors were detected. The skill relies exclusively on markdown-based instructions to guide agent behavior during text processing.- [PROMPT_INJECTION]: The skill is designed to process external consultation logs, creating a surface for indirect prompt injection. However, it incorporates strong counter-instructions for anonymization and validation.
- Ingestion points: User-provided documents (.md, .txt) or pasted content as defined in intake-checklist.md.
- Boundary markers: Absent; input text is processed as raw narrative data.
- Capability inventory: File system read and write operations used to generate reports (references/batch-processing-rules.md).
- Sanitization: Extensive de-identification rules for names, company entities, and contact information specified in output-template.md.- [COMMAND_EXECUTION]: The skill instructs the agent to read from and write to the local file system. Specifically, it generates reports in the same directory as the input documents using a standard naming convention. These operations are limited to text processing and report generation.
Audit Metadata