L1: Input & Validation
The foundational layer that validates all data entering your agent workflows, ensuring clean inputs for reliable processing.
What is Layer 1?
Layer 1 (L1) serves as the entry point for all agent workflows. Before any data reaches your agents, it passes through validation gates that check data types, required fields, format compliance, and business rules. This ensures only clean, verified data enters the processing pipeline.
Supported Input Types
- Text & JSON - Structured data with schema validation
- Files & Documents - PDFs, spreadsheets, and text files
- Images & Audio - Multi-modal inputs for vision and speech agents
- URLs & Web Content - Automatic content extraction and parsing
- API Payloads - REST and webhook data with authentication
- Form Submissions - User inputs from custom forms and interfaces
Validation Features
Schema Validation
Define custom schemas with required fields, data types, and format patterns. Invalid inputs are rejected with detailed error messages.
Integrity Checking
Automatic checksums and deduplication ensure data hasn't been corrupted or duplicated during transmission.
Business Rules
Apply custom business logic constraints—value ranges, conditional requirements, and cross-field validations.
Error Handling
When validation fails, L1 provides detailed error responses with field-level feedback. Configure error handling strategies: reject invalid data, apply defaults, or route to manual review queues.