writing-fragments
Pass
Audited by Gen Agent Trust Hub on May 19, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The logic of the skill is consistent with its primary purpose of helping users ideate and document writing fragments. It does not contain any hidden scripts, unauthorized network operations, or hardcoded credentials.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its operational design.\n
- Ingestion points: The agent is explicitly instructed to re-read the entire markdown file from the disk before every write operation to preserve user edits (SKILL.md).\n
- Boundary markers: There are no instructions to use delimiters or specific directives to ignore instructional language that might be present within the file content.\n
- Capability inventory: The skill allows the agent to perform file system read and append/write operations based on a user-provided path.\n
- Sanitization: There is no evidence of sanitization, filtering, or validation applied to the content read from the file before it is incorporated into the agent's context.
Audit Metadata