grant-writing

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and prioritize content from external sources, which creates a surface for indirect prompt injection attacks. Any grant solicitation or example document provided by a user could potentially contain instructions that override the agent's behavior.
  • Ingestion points: The agent is instructed to read external grant solicitations, evaluation criteria, and user-supplied example proposal directories as specified in SKILL.md and references/example-proposal-patterns.md.
  • Boundary markers: The skill lacks explicit delimiters or instructions to treat external text as untrusted data. Furthermore, it specifically directs the agent to "Treat the solicitation and scoring rubric as higher priority than this skill," which increases the risk of the agent following malicious instructions embedded in those documents.
  • Capability inventory: The skill primarily performs text analysis and generation, though it integrates with other writing workflows that may have additional capabilities.
  • Sanitization: No sanitization or validation of the external input is mandated before the agent processes and acts upon the information.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 12:50 AM