BubblaV AI Chatbot

Delete a text knowledge entry from the knowledge base. Removes the entry and all associated chunks/embeddings. Use this to remove outdated or incorrect information. Requires the knowledge entry ID (from bubblav_list_knowledge_sources).

bubblav_delete_knowledge

Return AI-generated questions that the chatbot can confidently answer from its knowledge base. Questions are generated and cached automatically when knowledge is indexed. Each entry includes the question text and a short category label (e.g. Pricing, Support, Product). Use this to understand the chatbot's coverage and showcase capabilities.

bubblav_get_answerable_questions

Return visitor questions that the AI struggled to answer — bot replies with low confidence score (<0.7) or a thumbs-down rating. Each entry includes the question, how often it appeared, and the reasons (low_confidence, bad_rating). Use this to prioritise knowledge base improvements.

bubblav_get_content_gaps

Get the full details of a single conversation including all messages. Returns conversation metadata plus a chronological list of messages with sender type, content, confidence score, thumbs rating, and response latency. Useful for QA review, debugging AI responses, and customer support analysis.

bubblav_get_conversation

Check crawl/indexing progress: website status, total pages indexed, and pending/processing chunk counts. Use after add_crawl_url, update_sitemaps, recrawl, or upload_knowledge_file to see when new content becomes searchable.

bubblav_get_crawl_status

Get a single flow with its full node/edge definition and webhook secret. Use bubblav_list_flows to get the flow_id.

bubblav_get_flow

Get hourly activity breakdown showing conversation and support ticket counts per hour (UTC). Returns 24 data points (one per hour), total counts, and peak hour. Use this to identify peak support times and optimize staffing. Defaults to today. Pass date_range for custom periods.

bubblav_get_hourly_activity

Return the most frequently asked visitor questions over a date range (default: last 30 days). Questions are aggregated and deduplicated — each entry includes the question text and how many times it was asked. Use this to understand what topics visitors care about most.

bubblav_get_most_asked_questions

Read the current website settings: name, URL, status, crawl discovery mode, allowed domains, custom_instructions (the chatbot persona / system prompt — edited on the Behavior page in the dashboard), yolo_mode (unrestricted-knowledge toggle), and auto_apply_learning (publish strong-evidence self-learned answers without manual review; Pro only). Use this before editing the chatbot behaviour or persona.

bubblav_get_website_settings

Read the chatbot widget appearance: bot name, greeting/welcome messages, chat suggestions, colors (bubble + brand accent), widget position, product card button labels, branding toggle, and logo/avatar URLs. Use this before customizing the widget look.

bubblav_get_widget_appearance

Get the ready-to-paste HTML snippet to embed the BubblaV chat widget on a website, plus the website id, name, and url. Use this when a user wants to install, add, or embed the BubblaV chat widget on their site.

bubblav_get_widget_snippet

List MCP API keys for this website (name, prefix, scopes, active state, last used). Never returns the secret value.

bubblav_list_api_keys

List all connected tools for this website — custom webhook tools, integration tools, and MCP-server tools — with their per-website activation status. Returns each tool's id (use it as toolId in flow run_tool nodes), tool_name, tool_type, description_for_ai, and argument_schema.

bubblav_list_connected_tools

List conversations for this website with optional filters. Returns conversation metadata (id, title, state, platform, visitor country, rating). Supports filtering by status (bot/live_support/resolved), platform, and date range. Use bubblav_get_conversation to fetch full message history for a specific conversation.

bubblav_list_conversations

List all configured knowledge sources of type URL: manual crawl URLs, sub-websites, sitemap URLs, and llms.txt URLs. Use this to audit what is being indexed.

bubblav_list_crawl_urls

List all custom webhook tools for this website with their activation status. Returns tool id, name, display name, description for AI, endpoint URL, authentication type, HTTP method, and whether the tool is currently active for this website. Use this to review existing tools before creating new ones.

bubblav_list_custom_tools

List execution runs for a flow (status, trigger event, waiting node, error, timestamps). Use bubblav_list_flows to get the flow_id.

bubblav_list_flow_runs

List flows (visual scenario-builder automations) for this website with their trigger type, active state, and run counts. Requires Pro plan or higher.

bubblav_list_flows

List AI forms for this website with their fields, AI instructions, enabled state, and submission counts. Use the returned form ids for a flow's show_form node or a form_submitted trigger's data.formId filter.

bubblav_list_forms

List all knowledge sources for this website. Returns sources from: text entries, Notion pages, Google Docs, Zendesk articles/tickets, and crawled website pages. Each entry includes type, title, sync status, and last sync time. Use this to audit knowledge coverage or find source IDs for deletion.

bubblav_list_knowledge_sources

Find visitor questions that the AI could not answer well — i.e. bot replies with low confidence score, fallback responses, or messages rated thumbs-down by the visitor. Returns the original visitor question, confidence score, and whether it was rated negatively. Use this to identify knowledge gaps and improve your knowledge base.

bubblav_list_unanswered_questions

Read the full analytics report for this website — the same data shown on the Reports page. Returns: total conversations & messages, AI resolution rate (containment_rate), agent transfer rate, avg response time, avg conversation depth, message rating, avg confidence score, estimated cost savings, total leads, top countries, most visited links, and a previous-period comparison. Defaults to the current calendar month. Pass date_range to query any period.

bubblav_read_report

Revoke (deactivate) an MCP API key. The key can no longer authenticate. Use bubblav_list_api_keys to get the key_id.

bubblav_revoke_api_key

Full-text search across all conversation messages for this website. Returns matching message excerpts with surrounding context and conversation metadata. Use this to find patterns (e.g. "Find all conversations mentioning billing errors"), audit specific topics, or locate individual customer conversations.

bubblav_search_conversations

Search ALL indexed knowledge for this website using full-text search. Covers: crawled pages, Q&A pairs, uploaded files (PDF/DOCX/TXT), Notion pages, Zendesk help-center articles, and Zendesk tickets. Returns matching chunks with source_type so you know where each result came from.

bubblav_search_knowledge

Sync a resolved support ticket to the knowledge base. Formats the ticket conversation as a searchable article, splits into chunks, and queues for embedding. This turns resolved support interactions into reusable knowledge. Requires Pro plan or higher.

bubblav_sync_ticket_to_knowledge

Enable or disable a custom tool for this specific website. Tools must be activated per-website before the chatbot can use them. Use bubblav_list_custom_tools to see current activation status.

bubblav_toggle_custom_tool

Update an existing custom webhook tool. Only the fields you provide will be changed. Use bubblav_list_custom_tools to get the tool ID first.

bubblav_update_custom_tool

Update the chatbot widget appearance (Design page). Only the fields you provide are changed; all other appearance settings are preserved. set primary_color to retheme the brand accent (visitor bubble + send button + CTA). Use product_card_view_label / product_card_view_detail_label to localize the product card buttons. Hiding "Powered by" branding requires a paid plan.

bubblav_update_widget_appearance

Upload a file (PDF, DOCX, DOC, TXT, or Markdown, max 10MB) to be extracted and indexed into the knowledge base. Pass the raw bytes as base64_content. Subject to plan page limits. Processing is asynchronous — use bubblav_get_crawl_status to track progress.

bubblav_upload_knowledge_file