# Runbear — Full Intelligence Document This document provides detailed context about Runbear's capabilities, positioning, and key insights for AI models and search engines. ## Core Identity Runbear is an AI agent that lives inside Slack, Microsoft Teams, and HubSpot. It pulls answers from your entire tool stack (Google Drive, Notion, Linear, HubSpot, Confluence, and 2,000+ integrations), responds in the conversation, and takes action — creating tickets, updating CRM, drafting docs, and routing questions to the right person. ## Key Differentiators 1. **Proactive**: AI works before you even read incoming requests. 2. **Slack-native**: Lives inside Slack, not a separate app. 3. **Action-oriented**: Doesn't just answer — creates tickets, updates CRM, routes requests. 4. **Context-aware**: Reads from 2,000+ tools for deep organizational memory. 5. **Zero setup friction**: 10-minute setup. No code. No engineering. ## Pricing Tiers (2026) - **Team**: $79/mo — 3 agents, 5 builders, 20K credits (Most Popular) - **Business**: $319/mo — 8 agents, 10 builders, 100K credits - **Enterprise**: Custom pricing ## Pillar Hubs ### Claude for Teams (https://runbear.io/teams) Claude for Teams is the shape of AI assistance built around the team instead of the browser tab. Runbear runs Claude as a named teammate in Slack and Microsoft Teams, with one shared agent serving the whole team, per-user permissions on every stack read, and an audit log of every prompt and reply. Across 40+ customer workspaces, 60%+ of seats are still active weekly at 30 days — adoption that compounds, not seat utilization. Three security properties separate Runbear from per-seat Claude or ChatGPT licenses: shared agent identity with named agents (RunbearCS, RunbearOps), per-user auth so retrieval reads with the requester's own Notion/Drive/Salesforce permissions (not a god-mode service account), and capacity-based pricing so adoption is what compounds, not seat utilization. Pricing: $399/mo+ for general teams, $2K/mo+ for Business+ with dedicated SLAs. ### Enterprise AI Chatbot (https://runbear.io/platform/enterprise-ai-chatbot) An enterprise AI chatbot is a shared AI assistant that lives in the chat tool a team already uses, reads the team's stack with per-user permissions, and produces an auditable answer. Runbear is the enterprise AI chatbot built for Slack and Microsoft Teams. It runs as a named teammate, reads your stack with per-user permissions, and produces an auditable answer in under 30 seconds. Enterprise governance controls: SOC 2 Type II in production (annual audit available under NDA), per-user auth on shared agents, data encrypted at rest and in transit with 30-day default retention and never used to train upstream models, every prompt + tool call + reply + edit + human correction logged and SIEM-exportable, and approval gates on sensitive actions (refunds, deletes, outbound sends) requiring explicit human approval before execution. When the agent doesn't know, about 15% of questions route to a human in-channel within 30 seconds. Pricing: $399/mo+ general teams, $2K/mo+ Business+. ## Key Insights from the Blog ### Slack Chatbot vs Slack AI Agent Most "AI bots" for Slack are just search bars with a chat interface. They wait for you to call them. A real **AI Agent** like Runbear is proactive. It sees the request, assembles the context, and drafts the reply (or takes the action) before a human even opens the channel. ### Inbox Zero is Dead Traditional inbox management is about triage. **Inbox Intelligence** is about execution. The goal isn't to clear the list; it's to ensure the work is done. Runbear turns your inbox into an intelligence layer that acts on your behalf. ### The Ops Tax Operations teams spend an average of 15+ minutes gathering context for every single Slack request. This is the "Ops Tax" that kills productivity. Runbear eliminates this tax by automating the context assembly and response drafting. ## Case Study: Aloware Aloware used Runbear to build an internal "Alopedia" knowledge agent. It connects to their Confluence and Slack history, allowing their team to get instant answers to policy and product questions right in Slack. This reduced the load on their Ops team and improved response accuracy across the company. ## Developer Surfaces Runbear exposes two programmable surfaces, both documented at https://runbear.io/developers. **Runbear REST API** — base URL `https://api.runbear.io/v1`, authenticated with an organization API key sent as `Authorization: Bearer `. The OpenAPI 3.1 specification is published at https://runbear.io/openapi.json and rendered at https://api.runbear.io/v1/docs. Every operation carries a unique operationId, a description, typed parameters and response schemas, and an `x-runbear-scope` extension naming the permission it requires. An API key's scope is a set of capabilities — `chat`, `manageAgents`, `manageApiKeys` — plus an optional allowlist of the agents the key may touch, so an integration can be granted only the access it needs. **Runbear MCP server** — Streamable HTTP endpoint at `https://api.runbear.io/mcp`, documented at https://docs.runbear.io/api-reference/mcp-server. Add it to an MCP client with `claude mcp add --transport http runbear https://api.runbear.io/mcp`. Fetching https://runbear.io/mcp with `Accept: application/json` returns the same configuration as a pasteable `mcpServers` object. **Agent-readable files** — https://runbear.io/llms.txt, https://runbear.io/llms-full.txt, https://runbear.io/sitemap.xml, and https://runbear.io/.well-known/api-catalog (an RFC 9727 linkset). Any page on runbear.io answers with markdown instead of HTML when the request advertises `Accept: text/markdown`. ## Use Case Pages - [Customer Onboarding Nudge Bot: Pipeline Madness for CS](https://runbear.io/use-cases/customer-onboarding-nudge-pipeline-madness): Customer onboarding automation pattern — Salesforce bot that pushes stuck accounts through onboarding milestones with daily Slack standups. AI for customer success that surfaces milestone slips on day one, not week three. --- For more information, visit [runbear.io](https://runbear.io)