manuscript-optimizer

Pass

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists entirely of natural language instructions for an AI agent to act as a manuscript optimizer. It establishes a logical framework for scientific reasoning, claim architecture, and structural revision without requesting or using any external tools, network access, or file system modifications.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data (user-provided research manuscripts).
  • Ingestion points: User-provided manuscript text passed to the agent during a session.
  • Boundary markers: None explicitly defined within the skill instructions to separate manuscript content from agent instructions.
  • Capability inventory: No tool use, subprocess calls, network operations, or file-write capabilities are requested in the frontmatter or body.
  • Sanitization: None present. Given the absence of actionable capabilities (tools), the risk of indirect prompt injection being used to cause harm is negligible.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 18, 2026, 04:19 PM
Security Audit — agent-trust-hub — manuscript-optimizer