plugin-agent-orchestrator

Overview

Hierarchical Multi-Agent orchestration for NocoBase AI Employees. Enables Leader agents to delegate tasks to Sub-Agent employees without modifying core plugin-ai.

Features

• Hierarchical Delegation: Allows AI Leader agents to break down complex tasks and assign them to specialized sub-agents.
• Seamless Integration: Plugs directly into the existing AI Employee framework.
• Execution Tracking: Monitor sub-agent task execution and responses within the main chat interface.
• External RAG Search: Exposes external_rag_search so leaders and sub-agents can retrieve context from NocoBase knowledge bases, including external RAG services.

Usage

1. Enable the plugin in the NocoBase Plugin Manager.
2. Navigate to the AI Employee configuration page.
3. Configure a primary "Leader" agent.
4. Add available "Sub-Agents" as tools or skills to the Leader agent.
5. Interact with the Leader agent; it will automatically delegate tasks when necessary.

External RAG / Embedding service

Recommended approach: keep source ownership in NocoBase and move chunking, embedding, vector storage, and retrieval to an external lightweight RAG service. This plugin should not own vector infrastructure directly. It delegates search to plugin-knowledge-base, which already supports knowledge bases of type EXTERNAL_RAG.

Architecture

1. NocoBase stores the source metadata:
   • uploaded file attachment id, filename, storage id, URL, owner/role access, or
   • datasource/collection/record id for database-backed knowledge.
2. The external RAG service owns:
   • document fetching or receiving source payloads,
   • parsing/chunking,
   • embedding,
   • vector index,
   • retrieval.
3. NocoBase creates a knowledge base with type = EXTERNAL_RAG and options such as:
   • ragProvider: "external-http"
   • ragApiUrl: "https://rag.example.com/search"
   • ragApiKey
   • ragNamespace
   • ragTopK
   • ragScoreThreshold
4. Agents use the external_rag_search tool. The tool calls plugin-knowledge-base.searchKnowledgeBases(), so access control and mixed local/external search remain centralized.

External search contract

The built-in external-http strategy expects:

POST /search
Authorization: Bearer <ragApiKey>
Content-Type: application/json

{
  "query": "search text",
  "topK": 5,
  "scoreThreshold": 0.3,
  "namespace": "optional-kb-namespace",
  "filter": {}
}

Response:

{
  "results": [
    {
      "id": "chunk-or-source-id",
      "content": "matched text",
      "score": 0.82,
      "metadata": {
        "fileId": "123",
        "filename": "contract.pdf",
        "collection": "orders",
        "recordId": "456",
        "sourceUrl": "/api/attachments/123:download"
      }
    }
  ]
}

The important rule is that every result should return enough metadata for NocoBase to resolve the original source: fileId/filename for files, or collection/recordId for datasource records.

Lightweight open-source service target

For a small but useful deployment, use a standalone service with:

• Qdrant or LanceDB for vector search.
• FastAPI or Express for /ingest, /delete, and /search.
• A small embedding model such as BAAI/bge-small-en-v1.5, intfloat/multilingual-e5-small, or another model matching the deployment language.
• A namespace per NocoBase knowledge base.

Minimum API surface:

• POST /ingest: receive { namespace, source, content | fileUrl | storageRef, metadata }.
• POST /delete: receive { namespace, sourceId }.
• POST /search: implement the contract above.

For NocoBase datasource knowledge, send each record as a source document with metadata { collection, recordId, fields, updatedAt }; the RAG service chunks and embeds the selected text fields, then returns collection and recordId in search results.

E5/OpenAI-compatible embedding

If the external RAG service uses an E5 family model behind an OpenAI-compatible /v1/embeddings API, configure the NocoBase knowledge base with ragProvider: "e5-http". plugin-knowledge-base will read the selected embedding LLM service and forward its baseURL, apiKey, and model to the RAG service.

• embed user queries as query: <question>;
• embed document chunks as passage: <chunk>;
• use the same model for ingest and search;
• recreate the vector collection when changing models or vector dimensions.

The RAG service still owns chunking, embedding calls, vector storage, and metadata mapping. NocoBase owns the LLM service configuration and search authorization path.