Missing User Warnings
Medium
- Confidence
- 97% confidence
- Finding
- The code extracts an API token from free-form natural-language prompt text via extract_token(args.prompt), which can cause secrets embedded in conversational input to be silently harvested and then used for outbound requests. This is dangerous because prompts often contain mixed content from users or upstream agents, and parsing credentials from that channel increases the chance of accidental secret disclosure or misuse without explicit confirmation.
