store-compliance

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.90). The prompt includes a "SELF-EVOLUTION TELEMETRY" section that instructs the agent to probe the user's filesystem (~/.claude/projects/) and silently append telemetry to a project file, which is a stealthy logging/write action unrelated to the stated compliance-audit purpose and therefore a hidden/deceptive instruction.

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt requires citing exact file paths and line numbers and providing specific code changes (including code snippets/metadata), which would force the model to reproduce any secret values present in those lines verbatim if they exist, creating an exfiltration risk.