MCP Clients / Overview
Connecting AI Clients
Telara's supported employee connectors use OAuth 2.1 with PKCE. Employees select the approved connector and sign in once; Telara applies their organization's access and action policies on every request.
How It Works
One connection for governed work
A supported platform presents Telara in its directory, marketplace, or approved enterprise catalog. The employee selects Telara, completes OAuth once, and works with the same governed tools in the assistant they already use.
Choose Telara from the AI platform’s approved directory, marketplace, or organization catalog.
Complete OAuth 2.1 with PKCE using your work identity. You never paste an API key, client secret, tenant ID, or MCP configuration.
Your platform receives the Telara tools permitted by your organization and policy. Ask it to investigate company context or perform approved work.
Any write actions (create issue, send message, etc.) are checked against your permission policies. Actions set to "require approval" pause for human review.
Coding Agents
IDE tools and coding agents
AI-powered IDEs and coding agents. Prefer the approved marketplace or organization deployment path for employee use rather than copying a server URL or credentials.
One command to connect all your AI tools to Telara. Handles login, config selection, and context switching across Claude Code, Cursor, and Windsurf.
Agentic CLI for software engineering tasks. Connects via a JSON config file in your project or home directory.
AI-first code editor built on VS Code. Add Telara through the MCP Servers settings panel or a JSON config file.
AI-powered IDE with Cascade agentic flow. Configure via the Windsurf settings or a config JSON file.
Agent Frameworks
Agent frameworks
Python and JavaScript frameworks for building custom AI agents. Connect Telara as a tool provider in a few lines of code using each framework's MCP adapter.
Python and JS framework for building LLM apps. Use the langchain-mcp-adapters package to load Telara tools.
Multi-agent conversation framework. Connect via SseMcpToolAdapter from the autogen_ext package.
Role-based multi-agent framework. Add Telara as an MCPServerSSE tool available to any agent in your crew.
Lightweight SDK for building agentic applications. Attach Telara as an MCPServerSse in your agent definition.
Transport
SSE transport
All supported clients use SSE (Server-Sent Events) to stream tool results from Telara. SSE is a persistent HTTP connection — your tool connects once, and Telara streams responses back in real time as your agent calls tools.