proposal-ideate

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches publication metadata and abstracts from well-known scholarly APIs, including Crossref, DBLP, and arXiv, using read-only GET requests.
  • [INDIRECT_PROMPT_INJECTION]: The skill handles untrusted external data from web fetches and user-supplied documents (e.g., topic lists or thesis chapters). It includes explicit instructions to treat this content as data rather than instructions and to ignore any embedded commands or formatting that attempts to override agent behavior.
  • [DATA_EXFILTRATION]: The skill reads and writes session state to local markdown files (<slug>.notes.md and <slug>.md) within the user's workspace, which is essential for its persistence functionality.
  • [COMMAND_EXECUTION]: The instructions refer to other skills within the same tool suite, such as proposal-lit-search and proposal-troubleshoot, to provide integrated functionality for proposal development.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 09:46 AM
Security Audit — agent-trust-hub — proposal-ideate