abstract-writing

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to analyze user-provided manuscripts, introductions, and results to draft abstracts. This creates a processing surface for indirect prompt injection where instructions hidden in the user-provided text could influence the agent's behavior.
  • Ingestion points: The agent reads and interprets external manuscript content and research notes provided by the user as described in SKILL.md.
  • Boundary markers: The instructions do not specify any delimiters or safety warnings to distinguish between the content to be analyzed and potential instructions embedded within that content.
  • Capability inventory: The skill is restricted to text analysis and drafting; it does not utilize any tools for network communication, system modification, or command execution.
  • Sanitization: No explicit sanitization or filtering of the user-provided input is performed.
  • [EXTERNAL_DOWNLOADS]: The reference file references/example-abstract-patterns.md contains links to research papers on arxiv.org. This is a well-known service for academic preprints, and the links are used as stylistic references to guide the abstract drafting process.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 12:50 AM