data-management-plan
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill acts as a template-based document generator that processes user-supplied research specifications into formatted grant sections. It operates entirely on local files using standard reading and writing tools and lacks any network capabilities or privileged operations.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection because it ingests external data from the file path provided to the --input argument.
- Ingestion points: Research specifications and grant drafts loaded from the path provided to --input in SKILL.md.
- Boundary markers: The instructions do not define specific delimiters to isolate the ingested text from the agent's procedural logic.
- Capability inventory: The skill employs file system tools (Read, Grep, Glob, Write) and internal agent coordination tools (Agent, Task).
- Sanitization: Input data is parsed for metadata extraction without explicit filtering or validation steps.
Audit Metadata