> For the complete documentation index, see [llms.txt](https://docs.stackai.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.stackai.com/stackai-mcp/overview.md).

# Overview

Use StackAI from a connected assistant to find a project, understand how it works, retrieve company knowledge, make a reviewed change, or inspect a failed run. You can start with a plain-language request; the assistant discovers and calls the required tools.

MCP, the Model Context Protocol, is the connection between your assistant and StackAI. StackAI checks your sign-in, active organization, and permissions for the requested resources.

## Start with one useful task

1. [Connect your assistant](/stackai-mcp/connect-and-choose-your-organization.md) and confirm the active organization.
2. Try [Explore my workspace](/stackai-mcp/starter-prompts/explore-my-workspace.md) or [Explain a workflow](/stackai-mcp/starter-prompts/explain-a-workflow.md). Neither prompt asks the assistant to run or edit a project.
3. Open a returned project link and check the result. Then choose a task below.

| I want to…                           | Start here                                                                                                         |
| ------------------------------------ | ------------------------------------------------------------------------------------------------------------------ |
| Find an answer in my source material | [Find an answer in company knowledge](/stackai-mcp/starter-prompts/find-an-answer-in-company-knowledge.md)         |
| Run a published workflow             | [Inspect and run a published project](/stackai-mcp/workflow-recipes/inspect-and-run-a-published-project.md)        |
| Make and review a workflow change    | [Edit, validate and publish deliberately](/stackai-mcp/workflow-recipes/edit-validate-and-publish-deliberately.md) |
| Reuse instructions in a workflow     | [Attach a skill to an LLM node](/stackai-mcp/workflow-recipes/attach-a-skill-to-an-llm-node.md)                    |
| Compare results before rollout       | [Evaluate a project before rollout](/stackai-mcp/workflow-recipes/evaluate-a-project-before-rollout.md)            |
| Understand a failed or slow run      | [Diagnose failures without rerunning](/stackai-mcp/starter-prompts/diagnose-failures-without-rerunning.md)         |

## Choose the right documentation

* **Starter prompts** are requests to paste into your assistant. The assistant selects the tools.
* **Workflow recipes** show a sequence of calls, with checks between steps.
* **Tool reference** describes each tool's arguments, results, and behavior.

The catalog contains 51 tools. Availability in your client depends on the connection and your permissions. A read-only connection does not expose create, edit, upload, execution, or evaluator-preview tools. It can still switch between organizations you belong to.

## What this connection covers

This MCP server lets an external assistant work with StackAI. It differs from adding an external MCP server as a tool inside a StackAI agent. Provider actions used inside workflows are also not individual tools in this MCP catalog.

Use the StackAI interface to authorize a new provider connection, create or edit triggers, or delete or export a project. Those operations are not exposed by these MCP tools.

## Before changing or running a resource

Creation attempts to publish the new project. Project editing saves a draft unless publication is requested. Running a project or generating evaluation results can also run its integrations. Updating a skill changes the version used by agents immediately.

State the organization, resource, intended change, and whether the assistant should save, publish, or run. Inspect uncertain results before repeating a write or execution request.

[Connection and recovery guidance](/stackai-mcp/best-practices-and-troubleshooting.md)


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.stackai.com/stackai-mcp/overview.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.