Top 200 ReposEXPLORER
Browse agents, categories and pages
Back to results
Planned, not doneThe source describes an intention, not a finished result.

Build a support chatbot that answers from your help docs with Dify

Turn help-centre pages into a Dify knowledge base, route questions with a classifier, answer from retrieved context and embed the bot on your website.

Try the promptRead the original (opens in a new tab)

The problem

How it works

Dify's tutorial builds a Chatflow support bot. Help documentation is imported as a knowledge base (local files, Notion, or a website sync through Jina or Firecrawl), chunked, embedded and searched with hybrid retrieval. A question classifier sends off-topic and community questions to fixed replies, while product questions go through knowledge retrieval into an LLM node that writes the answer.

Results and limitations

The tutorial describes a build, not a measured result. It recommends deleting bad chunks and running Dify's retrieval test before previewing the whole app, and notes that classification quality depends on a capable model and clearly separated categories.

Before you try it

Dify can be self-hosted with Docker Compose on at least 2 CPU cores and 4 GiB of RAM, or used as Dify Cloud; its licence is based on Apache 2.0 with extra conditions, so read it before commercial use. The website sync in the example is capped at 50 pages per knowledge base. Keep personal customer data out of the knowledge base, give users a route to a human, and do not let the bot decide refunds, billing or legal questions. Results depend on your setup and were not independently reproduced by Top 200 Repos.

Tools and repositories

Tools
Dify (self-hosted or Dify Cloud) · An LLM and an embedding model provider · Optional: Jina or Firecrawl for website sync
Prerequisites
Docker Compose for self-hosting, or a Dify Cloud account, plus API keys for the chosen model providers.

Try it yourself: Plan a grounded support bot in Dify

Fill the fields to adapt the prompt. Empty fields stay as visible [placeholders].

Preview
Help me set up a Dify Chatflow that answers customer questions about [Product or service] using only our help content at [Help content location]. First list the pages or files you would import and flag any that contain personal data, internal notes or prices that change often, so I can exclude them. Propose question categories for the classifier, including an off-topic category with a fixed reply and a category that hands the customer to a human at [Human hand-off contact]. Write the LLM system prompt so the bot answers only from retrieved context, says when it does not know, and never promises refunds, discounts or account changes. Then give me [Number of test questions] test questions with the answer you expect, so I can run the retrieval test before anything is published.

Run the retrieval test and a preview session before embedding the bot on a public page.

Related videos

About firecrawl/firecrawl

About firecrawl/firecrawl

About langgenius/dify

Independent creators. Listing a video implies no endorsement, sponsorship or partnership.

Sources

  1. Customer service bot tutorial (opens in a new tab)

    Project documentation · docs.dify.ai · Undated · Source checked 11 Nov 2026, 22:03 UTC

    What this source supports
    • Builds a customer service bot as a Chatflow backed by a knowledge base
    • Knowledge can come from local files, Notion or a website sync through Jina or Firecrawl
    • The website example is limited to 50 pages per knowledge base
    • A question classifier routes off-topic questions to a direct reply without an LLM
    • The published app can be embedded in a website, and knowledge bases managed through an API
  2. langgenius/dify README (opens in a new tab)

    Project documentation · github.com · Undated · Source checked 11 Nov 2026, 22:03 UTC

    What this source supports
    • Open-source platform with workflows, a RAG pipeline, agents and LLMOps
    • Self-hosting with Docker Compose needs at least 2 CPU cores and 4 GiB of RAM
    • Dify Cloud is a hosted option
    • Licensed under the Dify Open Source License, based on Apache 2.0 with additional conditions
Report a problem

Reported by source

Crawl a docs site into clean Markdown for a RAG index

Use Firecrawl to map and crawl a website, returning each page as Markdown or structured JSON ready to chunk and embed.

github.comfirecrawl/firecrawl
Reported by source

Build a cited Q&A knowledge base over your own documents

Self-host RAGFlow, upload PDFs, Office files and scans, and ask questions that come back with traceable source chunks.

github.cominfiniflow/ragflow
Reported by source

Build a self-hosted AI agent workflow with human approval

Use n8n's visual canvas to wire an LLM agent to your apps, add tools and memory, and pause for human approval before it acts.

github.comn8n-io/n8n
Planned, not done

Build a voice agent that answers and places phone calls with Pipecat

Pipecat is an open-source Python framework for real-time voice agents; its telephony support lets a bot take inbound calls or dial out through providers such as Twilio and Telnyx.

github.com
Reported by source

Build an on-call investigation agent from your own alert history

Goutham Veeramachaneni connected Claude Code to Grafana's gcx CLI, his code and runbooks, then trained its skill on two weeks of past pages for his services.

gouthamve.dev · 21 Apr 2026anthropics/claude-code
Reported by source

Fill out repetitive web forms with Skyvern's vision-based browser agent

Skyvern drives a browser with an LLM and computer vision to complete forms, such as quote requests or contact forms, on sites it has not seen before.

github.commicrosoft/playwright