publisher-voice

Pass

Audited by Gen Agent Trust Hub on Oct 19, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process user-provided text for revision, creating a surface where untrusted data could attempt to influence the agent's output behavior.
  • Ingestion points: The default_prompt in agents/openai.yaml indicates the agent will process provided text for revision based on the skill's instructions.
  • Boundary markers: The skill lacks explicit technical delimiters (such as XML tags or triple quotes) to wrap untrusted input, though the instructions provide conceptual guidance for the AI to distinguish between source facts and instructions.
  • Capability inventory: The skill defines no external tools, network access, file writing capabilities, or shell execution paths.
  • Sanitization: No input sanitization or filtering logic is present.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 19, 2026, 06:26 PM