proof-of-work

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill mandates capturing and presenting actual command outputs (including environment variables and config files) as evidence, which can force an agent to expose API keys, tokens, or passwords verbatim in its output.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The skill's Validation Protocols (see SKILL.md "Step 3: Check for Known Issues" and modules/validation-protocols.md "Bug/Issue Research Validation" which explicitly instruct agents to perform web searches and check GitHub issues, forums, and community sources and to record citations), require ingesting and acting on open/public, user-generated content that can materially influence recommendations and follow-up actions, so it exposes the agent to untrusted third‑party content that could enable indirect prompt injection.