Integration

Bring your OKRs into every AI tool

OKR Studio's MCP Server lets Claude Desktop, ChatGPT, GitHub Copilot, Cursor, and Glean query your objectives, key results, and check-ins in real time.

The protocol carries no memory between calls. Pacing is state — and state lives in the tool, not the pipe. OKR Studio does the cycle-aware analysis first, then hands your assistant the answer.

Updated for the finalized 2026-07-28 MCP specification
Use cases

Your OKRs, wherever you think

Stop copy-pasting progress updates. Ask your AI assistant instead.

Claude Desktop

Ask Claude about your OKR progress, blockers, and team alignment in natural language. Get instant answers without opening another tab.

ChatGPT

Query your current cycle status, key result health, and check-in history directly from ChatGPT conversations.

GitHub Copilot

Give Copilot real-time OKR context so it can prioritize implementation work based on what actually matters this quarter.

Cursor

Let Cursor understand your objectives when generating code — ship features tied to measurable outcomes, not guesswork.

Glean

Surface live objectives and key results alongside the rest of your company knowledge, so enterprise search answers "what are we actually working toward?" too.

Capabilities

What the MCP Server does

Live OKR data at the moment of the prompt — including latest check-ins, status changes, and team alignment.

Real-time OKR queries

Ask plain-language questions about progress, scores, and blockers. The MCP server returns live data — not cached snapshots.

Check-in from any AI tool

Submit progress updates and check-in notes without leaving your AI assistant. One prompt, done.

Team alignment visibility

See how your team's key results connect to company objectives — directly in your AI workflow.

Cycle-aware context

The protocol has no memory — every request arrives cold. Pacing is state, and that state lives in OKR Studio, not in the pipe. So your assistant gets which cycle is active, what's due for check-in, and which KRs are behind pace for this point in the cycle — already computed, not re-derived from a raw data dump.

Secure by default

Bearer-token authentication with an API key you generate — and revoke — in OKR Studio. Every request is independently scoped to your organization, with no shared session state, over HTTPS.

10-minute setup

Add the MCP server URL to your AI tool's config. Paste your API key once. Start querying immediately.

Setup

Connected in under 10 minutes

1
Copy your MCP server URL
Find it in Settings → Integrations → MCP Server in your OKR Studio account.
2
Add to your AI tool config
Paste the URL into Claude Desktop, ChatGPT, Copilot, Cursor, or Glean's MCP configuration.
3
Paste your API key
Generate a key in Settings → API Access. Your permissions carry over automatically, and you can revoke it any time.
4
Start querying
Ask natural-language questions about your OKRs. Get real-time answers instantly.

Frequently asked questions

What is an MCP Server?

MCP (Model Context Protocol) is an open standard that lets AI assistants query external tools and data sources in real-time. OKR Studio's MCP server gives any MCP-compatible AI tool live access to your OKR data — objectives, key results, check-ins, cycle status, and team alignment.

Which AI tools support MCP?

Claude Desktop, ChatGPT (via plugins), GitHub Copilot, Cursor, Windsurf, Glean, and any tool that supports the Model Context Protocol. The list grows monthly as more AI tools adopt the standard.

Does OKR Studio's MCP server track the latest spec?

Yes. The server runs the finalized 2026-07-28 Model Context Protocol on the stable v2.0.0 C# SDK — stateless by default, with no initialize handshake and cache-hinted tool listings. Clients still negotiating earlier protocol revisions keep working, so upgrading your AI tool on its own schedule never breaks the connection.

How long does setup take?

Under 10 minutes for most teams. Add the MCP server URL to your AI tool's configuration, paste an API key generated in Settings → API Access, and start querying. No code changes or infrastructure setup required.

What can I ask the MCP server?

Anything about your OKRs: 'What are my team's objectives this quarter?', 'Which key results are at risk?', 'Show me check-in history for Q2', 'What's our overall cycle progress?'. Natural language queries return structured, real-time data.

Is my OKR data secure?

Yes. The MCP server authenticates with a Bearer API key you generate in Settings → API Access and can revoke at any time. Every request is independently scoped to your organization — there is no shared session state to hijack — and all communication is encrypted via HTTPS. No data is cached outside your OKR Studio account.

Which plans include the MCP Server?

The MCP Server is included on Professional ($9/month) and Enterprise plans. Free plan users can upgrade to access MCP along with unlimited cycles, integrations, and full Delvyn Studio access.

Give your AI assistant OKR context

Start your 14-day free trial — MCP Server included on Professional plans.