MCP Server
The Model Context Protocol lets your AI assistant talk to Shipchart directly. Ask Claude about your revenue, create check-ins from Cursor, or run a weekly review without opening a browser.
Setup guides
Pick your editor and follow the guide to connect in under 5 minutes.
Connect via Claude.ai
Claude.ai and Claude Desktop use OAuth 2.1 for MCP. No manual bearer token is required.
- Open Claude.ai, then go to Settings → Connectors.
- Click Add custom connector.
- Enter your Shipchart MCP URL:
https://shipchart.app/mcp. - Click Add to start OAuth.
- Sign in to Shipchart and click Authorize on the consent screen.
- Return to Claude and start using Shipchart tools in any conversation.
Bearer token setup for Claude Code, Cursor, Windsurf, Codex, and other clients remains unchanged.
Read tools
Require a token with read ability.
| Tool | Description |
|---|---|
dashboard_get | Portfolio summary — monthly revenue, project counts, stale projects |
projects_list | List projects with stage filter, sort, tech stack, and issue tracker info |
projects_get | Full project detail with brief, metadata, tech stack, deployment, revenue, links, and check-ins |
projects_get_context | Developer-optimized context dump — everything an agent needs to start working on a project |
revenue_list | Revenue entries with date range and type filters |
checkins_list | Recent check-ins with mood and status |
ideas_list | Ideas with scores and sort options |
integrations_list | Connected integrations with sync status |
goals_list_by_project | Goals for a specific project with progress and status |
goals_list | All goals across the portfolio, grouped by scope |
goals_portfolio | Portfolio-level goals with auto-tracked progress |
expenses_list | Expenses by project or portfolio with monthly cost totals |
financials_summary | Full financial overview — revenue, expenses, and profit |
Write tools
Require a token with write ability (or sync for the sync tool).
| Tool | Description |
|---|---|
checkins_create | Record a check-in on a project with mood tracking |
revenue_create | Log a revenue entry (recurring or one-time) for a project |
projects_create | Create a new project in the portfolio |
projects_update | Update a project's name, description, brief, metadata, stage, or URL |
ideas_create | Capture an idea in the parking lot with optional scores |
ideas_promote | Promote an idea to a real project |
goals_create | Set a goal for a project or across the whole portfolio |
goals_update_progress | Update progress on a manual goal or mark it achieved |
expenses_create | Track a recurring or one-time expense for a project or portfolio |
integrations_sync | Trigger a manual sync for a connected integration |
Prompts
Pre-built workflows your AI assistant can run. The first six are built-in; agent:* prompts are dynamically generated from your AI agents.
| Prompt | Description |
|---|---|
weekly-review-prompt | Portfolio review — stale projects, revenue trends, suggested focus |
project-health-check-prompt | Deep-dive on a specific project — revenue, activity, recommendations |
quick-capture-prompt | Rapid check-in session across all active projects |
shipchart:init | Bootstrap a project from codebase context — find/create project, infer stage, and populate context fields |
shipchart:discover-context | Refresh context fields for an existing project from codebase analysis |
switch-project-prompt | Load full project context and get a briefing to start working on a project |
agent:{slug} | Dynamic per-agent prompt — one for each active AI agent you've created (e.g. agent:growth-advisor) |
Resources
MCP resources that provide context to AI assistants.
founder://dashboard— Portfolio summary snapshotfounder://projects/{id}— Project detail by IDfounder://projects/{id}/context— Full developer context for a project (tech stack, issue tracker, deployment, brief)