manuscript-ingest
Pass
Audited by Gen Agent Trust Hub on Mar 5, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by ingesting untrusted manuscript text or PDF content for processing.
- Ingestion points: The skill reads user-provided manuscript text or PDF data and stores it in output/PAPER.md.
- Boundary markers: No delimiters or isolation instructions are present to prevent downstream tools from interpreting manuscript text as commands.
- Capability inventory: The skill has the ability to write files to the local file system (output/PAPER.md).
- Sanitization: No filtering or validation of the ingested text is performed to mitigate potential instruction injection attacks.
Audit Metadata