FastGPT
by labring
Builds AI agents on a visual Flow canvas backed by hybrid-retrieval knowledge bases, workflow and RPA nodes, bidirectional MCP, and an OpenAPI REST API.
Skills
Knowledge Base Ingestion
Imports txt, md, html, pdf, docx, pptx, csv and xlsx files or URLs, then chunks them with manual, direct or QA-split segmentation.
Hybrid Retrieval Search
Searches multiple reusable knowledge bases with hybrid retrieval and reranking, and tests recall from a single-point search panel.
Visual Workflow Orchestration
Composes dialogue and plugin workflows on a visual Flow canvas, chaining agent skills, RPA nodes and user-interaction steps.
Bidirectional MCP
Publishes apps as MCP tools and consumes external MCP servers, so FastGPT acts as both an MCP client and an MCP server.
Call Chain Logging
Records complete call-chain logs and conversation history for annotation, citation editing and application evaluation runs.
Related Agents
Cherry Studio
Runs a Windows, Mac, and Linux desktop client for OpenAI, Gemini, Anthropic, Ollama, and LM Studio models, with 300+ pr…
Crawl4AI
Crawls and scrapes web pages into clean Markdown for RAG pipelines, with CSS, XPath, and LLM-driven structured extracti…
RAGFlow
Open-source RAG engine with deep document parsing (tables, charts, images) and an official MCP server. Deploy via Docke…
Onyx
Open-source enterprise AI search and assistant platform — connects to Slack, GitHub, Confluence, Drive and 40+ sources…