abstract-title-contribution-writer

Pass

Audited by Gen Agent Trust Hub on May 5, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns detected. The skill uses standard file system tools (Read, Write, Edit, Glob) and web tools (WebSearch, WebFetch) strictly within the context of drafting and revising academic manuscript components.
  • [SAFE]: The instructions focus entirely on academic integrity, specifically warning against overclaiming results ('Avoid "first", "novel", "general"... unless the evidence board supports them') and providing 'Overclaim Downgrades' in the reference files.
  • [SAFE]: The skill operates on local paper directories and interacts with LaTeX files (main.tex, paper.tex) which is consistent with its stated purpose of assisting academic researchers.
  • [SAFE]: No obfuscation, data exfiltration, or unauthorized command execution patterns were found.
Audit Metadata
Risk Level
SAFE
Analyzed
May 5, 2026, 12:39 PM