API Reference
The AUVY API is a REST API for invoking receptors and managing core resources. This reference focuses on endpoints you can drive with an API key or share token. All endpoints return JSON, use standard HTTP methods, and follow consistent patterns for authentication, pagination, and error handling. For a step-by-step walkthrough, see Integrate the API.Base URL and version
The API base URL is https://api.auvy.ai. The current API version is v1 (all paths are under/v1/). The SDK uses this by default; you can override baseUrl for self-hosted or local dev and optionally set apiVersion if a future version is introduced.
Authentication
This reference primarily uses:| Credential | Header | Use case |
|---|---|---|
| API Key | Authorization: Bearer ak_live_... | Server-to-server, CLI, scripts |
| Share token | ?t=... or body field | Public receptor and trace flows |
JWT-only workspace-management routes are intentionally excluded from this API-key-focused reference set.
- TypeScript SDK
- curl
Request & Response Format
All request bodies and responses use JSON. SetContent-Type: application/json for POST/PATCH/PUT requests.
Success response:
| Code | Status | Meaning |
|---|---|---|
AUTH_REQUIRED | 401 | Missing or invalid credentials |
FORBIDDEN | 403 | Insufficient permissions |
NOT_FOUND | 404 | Resource does not exist |
VALIDATION_ERROR | 400 | Invalid request data |
RATE_LIMIT | 429 | Too many requests (check retryAfter) |
Pagination
List endpoints acceptlimit and offset query parameters:
- TypeScript SDK
- curl
Rate Limiting
Requests are rate-limited per API key or user. Rate limit headers are included in every response:retryAfter seconds. The SDK handles this automatically with configurable retry options.
Streaming
Invoke a receptor withstream: true to receive tokens in real time:
- TypeScript SDK
- curl
OpenAPI Specification
Download OpenAPI Spec
OpenAPI 3.1 specification — import into Postman, Insomnia, or code generators.
Endpoints
Health
Liveness plus optional authenticated workspace and brain context.
Config
Active model metadata, embedding settings, and cost estimates.
Receptors
Invoke neurons and pathways. List, manage, and share receptors.
Public
Public share endpoints for receptors and traces.
Traces
Execution history and conversation contexts.
Jobs
Job status, results, and streaming.
Grep
Keyword search across all resource types.
Search
Semantic search across memories.
Tools
Native tool metadata and input schemas.
Neurons
AI workers with model config, prompts, and tool access.
Pathways
Visual workflows connecting neurons and execution steps.
Reflexes
Tool collections from integrations and built-in tools.
Stakeholder Map
Relationship graph snapshots, briefings, and saved versions.
Memories
Vector-backed knowledge stores with ingestion and search.
Recollections
Memory scopes attached to traces and neurons.
Usage
Usage statistics and analytics.
Activity
Workspace activity feed.