# The first agent framework built for AI-first teams.

> Mindset delivers AI by using AI. Your builders build agents with a coding agent, and your operators run them in production from their AI client. Both work through Mindset's MCP server on one platform, and people decide what goes live.

## [Build with a coding agent](https://docs.mindset.ai/docs/get-started)

Claude Code or Cursor creates the agent, connects your systems, writes the tests and puts it in your product, following the Mindset skills.

```
Turn SPEC.md into a Mindset agent in Sandbox.
```

## [Run them from an AI client](https://docs.mindset.ai/docs/check-what-happened)

Ask what an agent did, why a run failed or what it costs, then test and roll out the fix. The answers come from the same runs, costs and audit trail as the Studio.

```
What did the refund agent cost last week?
```

## How Mindset fits together

Builders and operators change the configuration. Wherever an agent runs, it reads that configuration, so a change needs no release of your code.

Builders work from a coding agent and operators from an AI client, both through the MCP server, or in the Studio. In Mindset, an agent is a prompt and a model, held to an order by a script when needed. It uses connections to outside systems, operations on them, functions and widgets. Every change is a version, and every execution is a run. Every piece sits on one security and compliance layer: the org sets its posture once, and every agent works under it, with the audit trail and, once an admin turns it on, personal data redaction. The platform runs in your own cloud (AWS, GCP or Azure) or is hosted by Mindset. Your app reaches agents through the SDK, your server through the API, and people through the Hub and schedules.

## The plumbing is already built

Every production agent needs all of this. With a general agent framework your team builds it before the first agent ships. In Mindset it's built and running, so your team's time goes on what agents do for the business.

- [Deploying agents](https://docs.mindset.ai/docs/how-agents-are-started-and-reached): Hub, SDK, API, schedules, webhooks and AI clients
- [Environments](https://docs.mindset.ai/docs/environments): Sandbox and Production, kept apart in one org
- [Versioned configuration](https://docs.mindset.ai/docs/versions-and-going-live): Activate, compare or roll back, with no code release
- [Scripts, functions and widgets](https://docs.mindset.ai/docs/scripts): Ordered phases, fixed steps, and forms and charts in chat
- [Connections and credentials](https://docs.mindset.ai/docs/connections-and-operations): Keys never reach the model or the coding agent
- [Testing before it goes live](https://docs.mindset.ai/docs/test-an-agents-behavior): An untested operation needs an admin to go live
- [Logging and cost](https://docs.mindset.ai/docs/check-what-happened): Every run, model call and tool call, with its cost
- [One compliance posture](https://docs.mindset.ai/docs/compliance): Admins set the rules once, and every agent follows them
- [Admin-only settings](https://docs.mindset.ai/docs/org-settings): Only an org admin can change the rules, never an API key
- [Operations only](https://docs.mindset.ai/docs/security-model): Agents reach your systems only through actions you name
- [Approvals](https://docs.mindset.ai/docs/approve-what-an-agent-can-do): Admins decide which writes agents can make
- [Personal data redaction](https://docs.mindset.ai/docs/audit-and-personal-data): Personal data replaced before a model provider sees it
- [Tamper-evident audit](https://docs.mindset.ai/docs/audit-and-personal-data): Every action recorded, and never edited or deleted
- [Compliance frameworks](https://docs.mindset.ai/docs/audit-and-personal-data): Evidence for GDPR, SOC 2, ISO 42001, the EU AI Act and more
- [Data residency](https://docs.mindset.ai/docs/security-model): Your org's data stays in the region chosen when it's created
- [Your own AWS account](https://docs.mindset.ai/docs/host-it-yourself): Host Mindset yourself when your data or contracts require it

## Start

Ten minutes, once per project.

1. In the Studio, open **Settings → Systems**, choose your coding agent and copy the prompt it gives you. It adds the `mindset-admin` MCP server for your environment.
2. Paste it into your coding agent, then sign in when it asks. There's no key to paste.
3. Give it the skills.

```
Install the Mindset skills from https://docs.mindset.ai/skills and read mindset-mcp first.
```

Skill: [mindset-mcp](https://docs.mindset.ai/skills/mindset-mcp/SKILL.md)

## Builders

- [Get started](https://docs.mindset.ai/docs/get-started): Takes you from a new org to an agent running in your product, with your coding agent doing the building through Mindset's MCP server.
- [Keep your agents in your code](https://docs.mindset.ai/docs/keep-your-agents-in-your-code): Keeps each agent's prompt, model, script, functions and widgets as files in your repo, and has your coding agent apply them to Mindset as versions.
- [Start with your coding agent](https://docs.mindset.ai/docs/work-from-claude-code): Connects Claude Code, Cursor or another coding agent to Mindset, so it builds, tests and changes your agents for you.
- [Work with the MCP server](https://docs.mindset.ai/docs/work-with-the-mcp-server): Has your coding agent read Mindset's current method before it builds, start from worked examples, and hand a person's decisions back to them.
- [Move an agent from another framework](https://docs.mindset.ai/docs/move-an-agent-from-another-framework): Rebuilds an agent written with the OpenAI Agents SDK, LangGraph, CrewAI or your own code as a Mindset agent, kept as files in your repo.
- [Turn a skill or spec into an agent](https://docs.mindset.ai/docs/turn-a-skill-or-spec-into-an-agent): Turns a Claude skill, a prompt file, a spec or an MCP server you already have into a Mindset agent and the resources it needs.
- [Choose what to build](https://docs.mindset.ai/docs/choose-what-to-build): Decides which part of an agent's work goes in its prompt, a script, a function, a widget or another agent.
- [Build an agent](https://docs.mindset.ai/docs/build-an-agent): Creates an agent with its prompt, model and resources, hands work between agents, and chooses where people can reach it.
- [Connect a system](https://docs.mindset.ai/docs/connect-a-system): Lets an agent read from or write to one of your systems, through operations you choose.
- [Post to Slack or send email](https://docs.mindset.ai/docs/post-to-slack-or-send-email): Lets an agent post to a Slack channel through your own Slack bot, or send email through your Postmark account.
- [Add a function](https://docs.mindset.ai/docs/build-a-function): Gives an agent a fixed pipeline of steps it calls as one tool, for the part of its work that must come out the same every time.
- [Show a widget](https://docs.mindset.ai/docs/build-a-widget): Gives an agent a table, card, form or chart it shows in a conversation, laid out the same way every time.
- [Make it follow steps](https://docs.mindset.ai/docs/write-a-script): Gives an agent ordered phases for the part of its work that must wait or happen in order, including waiting for a person's answer.
- [Test before it goes live](https://docs.mindset.ai/docs/test-before-it-goes-live): Runs the test parameters of an agent's script and functions, runs its behavior tests, and keeps a change staged until they pass.
- [Move it to Production](https://docs.mindset.ai/docs/move-it-to-production): Applies the same agent files to Production once Sandbox passes, with Production's own connections, credentials and API key.
- [Put an agent on your page](https://docs.mindset.ai/docs/put-an-agent-on-your-page): Runs a Mindset agent in your web app with one backend endpoint and the mindset-agent element, in plain HTML, React or Vue.
- [Use your own UI](https://docs.mindset.ai/docs/use-your-own-ui): Drives a Mindset agent from your own chat interface with the UI-less client, and draws its quick replies, cards and widgets.
- [Let the agent act on your page](https://docs.mindset.ai/docs/page-tools): Lets an agent call functions in your page, tells it what the user is looking at, and fills in values the model must never choose.
- [Layout and theming](https://docs.mindset.ai/docs/layout-and-theming): Sizes the mindset-agent element, puts it in a floating bubble or a side panel, and matches its colors and dark mode to your app.
- [Start it from your server or a schedule](https://docs.mindset.ai/docs/start-it-from-your-server): Starts an agent from your backend, on a schedule or from another product's webhook, with no person in a chat.

## Operators

- [See what happened](https://docs.mindset.ai/docs/check-what-happened): Finds out what an agent did, whether it ran as often as it should, and whether it uses what it was given.
- [Diagnose and improve](https://docs.mindset.ai/docs/diagnose-and-improve-an-agent): Works out why an agent didn't do what you expected, and stages a fix as a version you test before it goes live.
- [Test behavior](https://docs.mindset.ai/docs/test-an-agents-behavior): Proves every step of a script works, checks how an agent behaves across many runs, and compares a change before you make it.
- [See and control cost](https://docs.mindset.ai/docs/see-and-control-what-it-costs): Shows what agents spend by agent, function, model, provider and user, and brings it down without making them worse.
- [Change something that's live](https://docs.mindset.ai/docs/change-something-thats-already-live): Changes an agent people already use, tries it before they meet it, puts it live, and rolls back if it goes wrong.
- [Approve what agents can do](https://docs.mindset.ai/docs/approve-what-an-agent-can-do): Decides which write operations agents can call, approves or withdraws each one once, and has a person decide single actions in Slack.
- [Audit and personal data](https://docs.mindset.ai/docs/audit-and-personal-data): Shows an auditor what agents did and what protects them, and keeps personal data away from model providers.
- [Roll out Mindset](https://docs.mindset.ai/docs/roll-out-mindset): Sets up your org, signs people in, makes the day-one decisions and takes the first agent to production, then keeps the fleet in shape.
- [Use your agents from AI clients](https://docs.mindset.ai/docs/use-your-agents-from-claude-and-other-ai-clients): Calls your org's agents from Claude, Claude Code, Cursor or Copilot, and runs them on your own model provider keys.
- [Host it yourself](https://docs.mindset.ai/docs/host-it-yourself): Runs Mindset inside your own AWS account, operates it from Backstage, and keeps it current release by release.

## Leaders

- [Why Mindset](https://docs.mindset.ai/docs/why-mindset): Why a CTO or Chief Product Officer chooses Mindset, who it suits and who it doesn't.
- [Compliance](https://docs.mindset.ai/docs/compliance): One compliance posture, set by your admins, that every agent follows whoever built it and however it's set up.
- [Mindset for your security review](https://docs.mindset.ai/docs/mindset-for-your-security-review): The answers your security or procurement team will ask for, each linked to the page with the detail.

## Live demos

Working Mindset agents, each showing one feature. Open one and try it.

- [Trip planner](https://docs.mindset.ai/demos/trip-planner): Every feature in one trip planner. Plan a Lisbon trip, and the Features tab lights each feature as it fires.
- [What’s wrong with this one?](https://docs.mindset.ai/demos/relay): The agent knows what's on screen. Select a delayed shipment and ask about it: an agent with no tools answers from the page.
- [Ask your data](https://docs.mindset.ai/demos/ask-your-data): Answers drawn as a dashboard. Ask a sales question, and the dashboard charts what the agent's functions return.
- [Plain HTML counter](https://docs.mindset.ai/demos/ferrier): An agent on a plain HTML page. The page loads one script and one tag, then marks each line of its snippet as it runs.