draft-section

Pass

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns such as prompt injection, obfuscation, or remote code execution were detected. The skill logic is entirely consistent with its described purpose.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted local data from the manuscript project files.
  • Ingestion points: The skill reads content from 'index.qmd' and 'references.bib' as defined in SKILL.md.
  • Boundary markers: None are present in the drafting instructions to delimit untrusted file content.
  • Capability inventory: The skill has access to Bash, Read, Write, Edit, Glob, and Grep tools.
  • Sanitization: There is no explicit sanitization or filtering of the content read from files. While these factors create an attack surface for indirect prompt injection, this is expected behavior for a tool that drafts content based on existing project data.
Audit Metadata
Risk Level
SAFE
Analyzed
May 16, 2026, 10:10 AM
Security Audit — agent-trust-hub — draft-section