code-history

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

• Third-party content exposure detected (high risk: 0.80). The skill's SKILL.md Step 3 explicitly instructs using the gh pr list --search '<commit-sha>' command to fetch PR metadata from GitHub (user-generated, public third-party content) and to use PR titles/messages to determine intent, which the agent reads and can materially influence its analysis.