visual-explainer

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

• Third-party content exposure detected (high risk: 0.80). The skill's required workflows (e.g., prompts/diff-review and other data-gathering steps) explicitly call out fetching and ingesting external, user-authored GitHub PR/diff and commit data (e.g., "gh pr diff 42", "Read all changed files in full", "read commit messages and PR descriptions" in prompts/diff-review and README), which are untrusted third-party content the agent must read and which directly drive analysis, diagrams, and follow-up actions—creating a clear vector for indirect prompt injection.