novel-summarizer

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external novel text provided by the user, which presents a surface for indirect prompt injection attacks where malicious instructions could be embedded in the text to influence the agent's behavior.
  • Ingestion points: Processes arbitrary novel text inputs provided by the user for summarization (SKILL.md).
  • Boundary markers: The instructions include explicit constraints to "strictly follow the original text" and "not perform self-creation or adaptation" (SKILL.md), which serves as a basic boundary but does not prevent all injection attempts.
  • Capability inventory: The skill's environment is restricted to the Read tool and lacks network access, file writing, or command execution capabilities, significantly limiting the potential impact of an injection.
  • Sanitization: No explicit input sanitization or validation mechanisms are implemented to filter potentially malicious prompts within the story text.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 06:49 PM
Security Audit — agent-trust-hub — novel-summarizer