read-memories

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill's Python steps read and print snippets from local conversation logs (which may contain API keys, tokens, or passwords) for the agent to "absorb" into its context, so even though it says not to echo raw logs, the LLM is exposed to secret values and could leak them — creating a high exfiltration risk.