fix-unnatural-line-breaks
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill implements a focused utility for text formatting. The included Node.js script (
scripts/lint.mjs) is self-contained, uses only built-in modules, and performs local regex-based analysis without any network activity or dangerous system calls. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external, untrusted content such as README files and source code comments, which serves as a potential ingestion point for malicious instructions. However, the risk is minimal as the skill's instructions specifically limit the agent to a narrowly defined text transformation (merging line breaks) and explicitly forbid broader content changes or restructuring.
- Ingestion points: The agent reads target files directly, and the
scripts/lint.mjsutility reads them vianode:fsduring detection. - Boundary markers: The skill instructions identify Markdown structural elements (headings, code blocks, lists) as boundaries to respect and ignore.
- Capability inventory: The skill is restricted to local text processing; it lacks network access, file-writing capabilities in the scripts, or dynamic code execution of external data.
- Sanitization: Not applicable, as the skill performs structural formatting rather than content evaluation.
Audit Metadata