continuous-learning

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill explicitly logs "full execution trace" and passes raw example sessions (via json.dumps) into the model prompt for skill generation, which can include API keys/passwords or tokens and thus could cause the LLM to see and verbatim-echo secrets in generated SKILL.md outputs.