agents

Inkeep Agents provides two interoperable ways to create AI agents: a drag-and-drop no-code visual builder for non-technical teams and a TypeScript SDK for engineering teams. Both approaches stay in full two-way sync, allowing technical and non-technical collaborators to edit and manage agents in the same platform. The platform supports real-time AI chat assistants for customer experience, help centers, technical documentation, and internal copilots. It also enables AI workflow automation such as updating knowledge bases, documentation, blogs, CRMs, and helpdesk triage. Key components include an agents API for configuration and execution, a visual builder UI, the @inkeep/agents-sdk for declarative agent definitions, a CLI with push/pull sync commands, and a chat UI component library. The framework supports multi-agent architecture, MCP tools with credential management, triggering via MCP, A2A, and Vercel SDK APIs, observability through traces and OpenTelemetry, and deployment with Vercel or Docker. Under the hood it uses the Vercel AI SDK for LLM provider integration. The project is licensed under Elastic License 2.0 with Inkeep Supplemental Terms, a source-available fair-code license.