powershell-code-review

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill asks the agent to accept inline-pasted PowerShell code and produce a markdown report listing security/credential issues without any instruction to redact secrets, which means the LLM may need to reproduce secret values verbatim from the submitted code in its output.

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

• Third-party content exposure detected (high risk: 0.70). The required runtime workflow accepts “GitHub URL” and references an input that can be fetched from an outsider source (public web/GitHub content) and then ingested as readable code text into the LLM context for review, enabling indirect prompt injection via that outsider-provided free text.