writing-beats

Pass

Audited by Gen Agent Trust Hub on Jul 17, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill instructions define a legitimate and transparent interactive writing workflow without any hidden logic or malicious intent.- [NO_CODE]: The skill consists entirely of markdown instructions for the AI agent and does not include any executable scripts, binaries, or external code dependencies.- [SAFE]: File system interactions (reading raw material and writing/re-reading the article) are performed at the user's direction and within the context of the specific task.- [PROMPT_INJECTION]: The skill was evaluated for indirect prompt injection risks. \n
  • Ingestion points: The agent reads raw material from user-provided markdown files (SKILL.md). \n
  • Boundary markers: The instructions rely on natural language context rather than explicit delimiters, which is typical for creative writing skills. \n
  • Capability inventory: The skill has file read and write capabilities. \n
  • Sanitization: No sanitization is performed on raw material, but the iterative, human-in-the-loop nature of the beat-by-beat process serves as a primary control against automated instruction following. The risk is assessed as safe.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 17, 2026, 02:40 AM
Security Audit — agent-trust-hub — writing-beats