Custom Agents
Integrate Vectora with custom agents in Python, Node.js, Go, or any language that supports HTTP.
Integration Options
| Method | Use Case | Complexity |
|---|---|---|
| MCP | Agent runs locally in Go/Rust | [x] |
| REST API | Remote agent in Python/Node/Go | [x] |
| ACP (Beta) | Distributed inter-agent communication | Beta |
REST API (Recommended to Start)
Vectora exposes a REST API (beta) that any agent can call:
curl -X POST https://vectora.app/api/search \
-H "Authorization: Bearer YOUR_TOKEN" \
-H "Content-Type: application/json" \
-d '{
"query": "How to validate JWT tokens?",
"namespace": "your-project",
"top_k": 5
}'MCP for Custom Agents
If your agent is written in Go/Rust, you can use MCP directly:
Your Agent (Python/Node) ↔ MCP Client ↔ Vectora MCP ServerSee MCP Integration for details.
Get Started
- Obtain an API token at console.vectora.app
- Configure your agent to call
/api/search - Test locally with
curl - Integrate into your workflow
Full Reference
- REST API Reference (beta):
/api/docs - MCP Spec: Model Context Protocol
- ACP (Beta): GitHub Discussions
Need a specific integration? Open an Issue
External Linking
| Concept | Resource | Link |
|---|---|---|
| MCP | Model Context Protocol Specification | modelcontextprotocol.io/specification |
| MCP Go SDK | Go SDK for MCP (mark3labs) | github.com/mark3labs/mcp-go |
| JWT | RFC 7519: JSON Web Token Standard | datatracker.ietf.org/doc/html/rfc7519 |
Part of the Vectora ecosystem · Open Source (MIT) · Contributors