assignment-composition
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external files such as
phase-1-dimensions.mdandgrain-context.mdto compose assignments. While this presents a surface for indirect prompt injection if these files contain malicious instructions, the structured nature of the output (a 9-section envelope) and the absence of high-risk capabilities like network access or arbitrary code execution minimize potential impact. - Ingestion points:
phase-1-dimensions.md,grain-context.md,references/handoff-envelope-spec.md,references/assignment-anatomy.md. - Capability inventory: Read, Write, Edit, Glob, Grep.
- Boundary markers: The skill instructs the agent to use blockquotes for the originating grain and structured headers for individual sections.
- Sanitization: No explicit sanitization or filtering of input content is specified.
Audit Metadata