writing-fragments
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill incorporates a mechanism to re-read the target markdown file from disk to preserve manual user edits. This creates a theoretical surface where untrusted data could influence agent behavior.
- Ingestion points: Target markdown file contents (SKILL.md).
- Boundary markers: Absent; the agent reads the file to maintain synchronization without specific delimiters.
- Capability inventory: Local file system read and write access.
- Sanitization: None; the agent is instructed to preserve changes exactly as they appear in the file.
Audit Metadata