Read and update issues, accounts, and contacts directly from AI tools that support the Model Context Protocol, including Claude, ChatGPT, and Cursor.
Every action runs on behalf of the authenticated user through OAuth, so an AI tool can only see and change the same data that the user can already access in the Pylon dashboard.
Point a tool like Claude Code at a Pylon link and have it investigate the issue, pull the full message history, and open a PR for the fix all in one flow.
Feed issues into an AI tool to generate a product spec, or ask questions about a customer in ChatGPT while pulling in Pylon data alongside other sources like Notion and Linear.
Grant the MCP access role to the Member or Admin users who need it. Access follows the same scoping as the Pylon dashboard — the MCP can never return data or make writes a user couldn't already perform in the UI.
In your Pylon dashboard, go to Settings → AI Controls → MCP Server to configure and enable the server.
Add the server URL https://mcp.usepylon.com in your AI tool of choice. The tool authenticates via OAuth 2.0, and connects over stateless streamable HTTP.
MCP is an open standard that lets AI tools like Claude, ChatGPT, and Cursor connect securely to external platforms and data. Pylon's MCP server exposes your Pylon data, including issues, accounts, contacts, and conversations, to any MCP-compatible AI tool, so your team can query and act on support data using natural language from inside the AI tools they already use.
Pylon's MCP server works with any MCP-compatible client, including Claude Desktop, Claude Code, ChatGPT, and Cursor. Once connected, these tools can read and update Pylon issues, search accounts, fetch contacts, and surface customer context, all via natural language.
Connected AI agents can search and update issues, look up account details, fetch contact records, add internal notes, and assign tickets. All actions are scoped to the same data and permissions as the authenticated user's Pylon dashboard, so nothing outside their normal access can be read or modified.
Yes. The Pylon MCP server uses OAuth 2.0 authentication, so AI tools connect on behalf of the authenticated user with session-scoped tokens. Access is strictly limited to data the user can already see in Pylon, and no extra permissions are granted beyond what exists in their normal account.
Enable the MCP server from your Pylon dashboard under Settings, then AI Controls, then MCP Server. Copy the MCP URL and paste it into your AI tool's MCP configuration. Authentication is handled via OAuth, so no API keys or custom infrastructure are required. Member or Admin seat access is needed to connect.
Pylon connects with the tools your team lives in — Slack, email, CRMs, ticketing systems,and more. Meet your customers where they are and streamline every support workflow.
Receive and respond to customer messages from Telegram
Manage B2B support directly in Slack channels alongside your team.
Connect Pylon to Granola to ingest call recordings
Sync Outlook Calendar meetings to customer records in Pylon
Automatically capture and log Google Meet call notes to Pylon
Automcatically link active incidents to affected customer conversations.
Enrich customer records with usage data from Amazon Redshift.
Create Shortcut stories from customer conversations automatically
Trigger PagerDuty incidents from critical customer-reported issues.
