AI tools & MCP

Give your AI the customer context behind the task.

Connect your AI tool to Outcomet so it can read customer feedback and look up the product context behind the work.

Connect your AI tool See an example
Before the code

The ticket is a starting point.

“Improve onboarding” names a task. The customer's own words help you ask a more specific question before deciding what to change.

Your task

Improve onboarding.

Customer feedback

“What do I do after setup?”

A question to investigate

What happens between setup and the first useful action?

A quote can point to a question. Check the surrounding evidence before treating it as a root cause or choosing a solution.

Customer context. Before the code.Watch on YouTube

Start with a product question.

01

Read the customer's words

Ask your tool to read recent feedback and its available source evidence. Check the customer wording before turning a request into a solution.

“Show me the latest 10 feedback items with the available customer wording and source references. Say when a source is missing.”
02

Find the problem behind a task

Search the problems your team has recorded. Use them to understand the obstacle you are trying to address and the questions that remain.

“Find problems about onboarding. Summarize the recorded problems and say what is still unclear.”
03

Look up the initiative

Find the relevant initiative, then ask for its linked feedback and product context where available. Missing connections remain questions for your team.

“Find initiatives about onboarding. For the relevant one, show its linked feedback and relationships. Do not infer links that are missing.”

Your workspace. Your AI tool.

Use an Outcomet account with workspace access and permission to connect an external client. The setup guide covers Claude, Codex and Antigravity.

Open the setup guide
  1. 01Choose your compatible AI client.
  2. 02Add Outcomet and sign in through your browser.
  3. 03Review and approve the requested workspace access.
  4. 04Ask a question and check the sources in the response.

Review the access you grant.

The connection uses browser sign-in through OAuth. Available actions depend on your role and approved access. Some tools can write data, including feedback, contacts and companies. Start with a read-only question and review any action that would change your workspace.

FAQ

Common Questions

What is MCP?
The Model Context Protocol lets compatible AI tools use tools and context from another service. Outcomet's MCP connection gives your client access to authorized product context in your workspace.
What do I need to connect?
You need an Outcomet account with workspace access, permission to connect an external AI tool, and an MCP client that supports browser sign-in through OAuth or a compatible proxy. You review the requested access when connecting.
Which AI tools can I use?
The setup guide covers Claude Code, Claude apps, Codex and Antigravity. Connection steps differ by client; access to custom connectors can also depend on your client's plan or administrator settings.
Can my AI tool change data?
Yes, when the relevant tools and permissions are available. Outcomet supports creating feedback and, with contact write access, creating or updating contacts and companies. Available actions depend on your role and approved access. Review the action before asking your tool to make changes.
How is this different from Nova?
Nova assists you inside Outcomet. MCP lets a compatible external AI client retrieve Outcomet context and use its available tools. The external client decides how to use that context; the two do not have identical capabilities. People still review evidence and make product decisions.

Bring the customer's words to your next task.

Connect your AI tool