proposal-writer

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE] (SAFE): The skill is composed of purely instructional markdown documents. No code or scripts were detected in the analyzed files.
  • [DATA_EXFILTRATION] (SAFE): No network requests (curl, wget, fetch) or data exfiltration patterns were found. Examples provided in the documentation use placeholder business data (e.g., 'TechCorp', '$84,000') rather than sensitive credentials.
  • [REMOTE_CODE_EXECUTION] (SAFE): There are no package manifests (package.json, requirements.txt) or commands that download and execute remote scripts.
  • [PROMPT_INJECTION] (SAFE): The instructions provide domain-specific frameworks (e.g., 'CLOSE Framework', 'Proposal Pyramid') without attempting to override the agent's core safety instructions or extract system prompts.
  • [CREDENTIALS_UNSAFE] (SAFE): No real API keys, tokens, or private secrets were discovered in the documentation templates.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 05:26 PM