elite-review

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill requires quoting exact files, line numbers, and code snippets from diffs (and to run CLI fetches like gh pr diff), so if those diffs contain API keys/passwords/cookies the model would be expected to reproduce them verbatim and no redaction guidance is given.

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

• Third-party content exposure detected (high risk: 0.70). The skill's Phase 1 "Gather the diff" workflow (e.g., "If PR number/URL given → fetch with gh pr diff" and "Read the full diff — every changed file") requires fetching and ingesting PRs/diffs from GitHub (user-generated, third‑party content) that the agent will read and use to drive review decisions, exposing it to untrusted external content.