short-drama-novel-analyze

Pass

Audited by Gen Agent Trust Hub on Aug 16, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted fictional text provided by users. An attacker could embed malicious instructions within a novel to attempt to influence the agent's analysis or override instructions during the extraction phase.
  • Ingestion points: External novel files read and processed via the scripts/novel_index.py script.
  • Boundary markers: The skill implements a robust chapter indexing system that binds specific content spans to SHA-256 hashes to ensure data integrity during processing.
  • Capability inventory: The skill executes local scripts to index text and aggregates these findings into multiple markdown and JSON documents within the project directory.
  • Sanitization: Instructions (NVA-04, NVA-06) explicitly require that analysis records use de-quoted function summaries and observable facts rather than copying source text, which helps prevent the execution or propagation of instructions hidden in the source material.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 16, 2026, 05:31 PM
Security Audit — agent-trust-hub — short-drama-novel-analyze