graph-analysis

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly ingests public X/Twitter follower data as part of its workflow (see SKILL.md "Build a network graph" POST /api/graph/build and description "API-powered network graph analysis for X/Twitter follower relationships"), which is untrusted user-generated content that the agent reads and uses to drive analyses and actions (influencer/bridge detection), meeting the conditions for potential indirect prompt injection.