The repository shows what the code should do. The customer timeline shows what actually happened. Connect Pathbound and Claude Code can use both while it reproduces a bug, traces an account issue, or builds a customer workflow.
Pathbound becomes a remote MCP server in Claude Code. The tools sit beside the repository context, so a technical task can use current customer evidence without copying production records into a prompt.
The resolved contact and company record, plus recent events, support conversations, meetings, and email history.
One HTTP endpoint registered by the Claude Code CLI, with browser OAuth for the user running the session.
Server-side filters and aggregation for narrow, token-efficient reads during debugging and analysis.
Project, user, or local MCP scope, depending on whether the Pathbound connection should follow one repository or the engineer’s Claude Code setup.
Register Pathbound as a remote HTTP MCP server, then complete OAuth from inside Claude Code. No Pathbound key needs to be pasted into an interactive session.
Add Pathbound from the terminal. The default scope is local; pass --scope user or --scope project when the connection should live at that scope.
Start Claude Code and run /mcp.
Select Pathbound and complete the browser sign-in when Claude Code requests authentication.
Ask Claude Code to read the relevant Pathbound contact or timeline before it diagnoses the customer issue.
claude mcp add --transport http pathbound https://mcp.pathbound.ai/mcpOnce connected, these are real jobs Claude Code runs against your customer data. Each recipe is one tested prompt—paste it and go.
When a support ticket comes in, look the customer up in Pathbound before drafting a reply. Pull their unified profile, read the prior conversations we’ve had with them across every channel, and check what they did in the product and on the site leading up to this. Summarize it in three lines—who they are and their plan, the relevant history, what they most likely need—then draft a reply that already accounts for all of it, so I never have to ask them to explain again.See the full recipe—Full-context support
More jobs in use cases, each with a tested prompt.
Run claude mcp add --transport http pathbound https://mcp.pathbound.ai/mcp, then open /mcp in Claude Code and complete the browser OAuth flow.
Use the default local scope while testing. Use --scope user when Pathbound should be available across your own projects, or --scope project when the repository should carry the shared server definition. Do not commit credentials.
Not for an interactive Claude Code session. Claude Code completes OAuth in the browser. Use a scoped REST API key only for a headless process that cannot open a browser.
It can read and update permitted customer records through Pathbound while you work in the terminal, but it does not replace the CRM. Pathbound resolves the CRM record with support, email, and website history underneath the session.
Only through write tools allowed by the Pathbound credential and connected source. Keep investigation and debugging read-only unless the task explicitly requires a change. Versioning and rollback are not live yet.
Use Claude Code when the repository or terminal is part of the job. Use Cowork for a repeatable operational queue or deliverable that does not require code.
Sign up, connect one CRM or drop the events snippet, and point your MCP client at mcp.pathbound.ai/mcp.