agent-execution-mode

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly requires pulling PR diffs and stated intent via GitHub MCP (e.g., references/WORKFLOWS.md and multiple pr-review rules) and using Figma MCP for designs (SKILL.md and references/WORKFLOWS.md), meaning it ingests user-generated third-party content (PRs, Figma specs) that the agent must read and that can materially influence decisions and tool actions.