monitor-ci

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

• Third-party content exposure detected (medium risk: 0.30). In the main loop, the orchestrator repeatedly calls the MCP tool ci_information with outsider-influenced CI state fields (select like cipeStatus, cipeUrl, ... failureClassification, hints, confidenceReasoning, taskOutputSummary etc.), then passes the resulting free-text fields into scripts/ci-poll-decide.mjs and the heavy-fetch paths read suggestedFixDescription/taskFailureSummaries/selfHealingSkipMessage for actionable decisions.