assembler
Pass
Audited by Gen Agent Trust Hub on Mar 1, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns or security vulnerabilities were detected. The skill's operations are limited to text processing and local file management within the expected project directory.
- [PROMPT_INJECTION]: Analysis identified an indirect prompt injection surface due to the ingestion of multiple external draft and note files without explicit boundary markers. This is assessed as safe because the skill's capabilities are restricted to document formatting and writing to manuscript.md, with no exposure to high-risk tools or network access.
- Ingestion points: scope.md, drafts/.md, notes/reviewer-feedback.md, notes/statistical-review.md, notes/ethics-summary.md, notes/papers/.md, notes/bibliography.md
- Boundary markers: None identified in the prompt templates.
- Capability inventory: File system read access for project files and write access for the output manuscript.
- Sanitization: Input content is concatenated directly without escaping or validation.
Audit Metadata