Wyregraf Documentation
Setup 2
Guide 20
- Agents & Networks (Pro)
- Chat Appearance and the Floating Launcher
- Citations & Answer Verification
- Coming from AI Engine or AI Puffer
- Content Enrichment & Chunking
- Conversations, Analytics, and Knowledge Gaps
- Creating a Chat
- Creating Embeddings
- Displaying the Chat
- Embeddings & Vector Profiles
- Getting Started with Wyregraf
- Hybrid Search & Reranking
- Improving Answer Accuracy
- Indexing PDFs, URLs & Code (Pro)
- Installation & Requirements
- LLM Profiles
- Summaries & Questions
- The Retrieval Lab
- Usage and Rate Limits
- What Visitors Can Do in the Chat
Troubleshooting 9
- My source says “Reindex required” — what do I do?
- The chat answers “no results” to questions my content covers
- The chat replies with my fallback message instead of an answer
- The chat returns an error instead of an answer
- The chat widget doesn’t appear on the page
- The embeddings queue looks stuck
- URL Scrape says my URL is “not allowed”
- Visitors see “You are sending messages too quickly”
- What should I include in a support request?
Reference 5
License 2
Hooks Reference 1
General 10
- Are visitor conversations saved, and for how long?
- How are my API keys stored?
- Is my content sent to the AI provider?
- What do the accuracy features cost to run?
- What does indexing and chatting cost?
- What happens to my data if I deactivate or delete Wyregraf?
- What’s the difference between the Retrieval and Chat types?
- Where are my embeddings and vectors stored?
- Which accuracy features should I turn on first?
- Which AI providers does Wyregraf support?