webhook-transforms

Webhook transformations convert incoming JSON payloads into sanitized markdown brain pages with structured metadata like slugs, types, and citations. The process applies deterministic functions that strip HTML tags to prevent XSS and ensure safe content ingestion.

Input sanitization for knowledge graph ingestion strips HTML tags and prevents raw scripts from reaching brain pages, reducing XSS risk. This safety-first approach ensures external event data like SMS or social mentions becomes secure, structured content.

Failed webhook event ingestion preserves raw payloads via a dead-letter queue and retries once before surfacing error types to the agent. This resilient pipeline prevents data loss during transformation failures.

Entity extraction and timeline updates run automatically on every transformed webhook event. The process extracts mentioned entities from incoming payloads, adds timeline entries, and syncs the knowledge graph to maintain chronological activity records.

Webhook transformation supports diverse ingestion scenarios including SMS received, completed meetings, and social mentions. Each external event type undergoes the same deterministic transformation to create sanitized brain pages with proper citations.

Raw HTML tags must be stripped during webhook data transformation to prevent XSS risk from reaching brain pages. This safety-first input sanitization ensures external payloads become secure, ingestible markdown content without script vulnerabilities.