> For the complete documentation index, see [llms.txt](https://learning.contextqa.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://learning.contextqa.com/enterprise-application-testing/ship/integrations-and-mcp.md).
# Connect Ship Integrations and MCP
Connect Ship to code hosts, issue trackers, CI/CD, private test targets, and AI agents while preserving least privilege and reviewable evidence.
> **Documentation owner:** [Deep Barot](https://contextqa.com/academy/) · **Last updated:** November 11, 2026
## Quick answer
Connect only the systems required for the workflow you are evaluating. Code providers supply authorized repository context, issue trackers preserve findings, CI/CD starts repeatable quality gates, tunnels reach approved private targets, and MCP lets supported AI clients retrieve or initiate allowed ContextQA work.
## Key takeaways
* Connect only the systems required for the workflow you are evaluating.
* Open Settings → MCP, then choose ChatGPT, Claude, Claude Code, Cursor, Codex, or the generic method shown in your workspace.
* Open a validated test plan and select CI / CD.
{% hint style="info" %}
**Watch the MCP setup:** [Connect Claude to ContextQA test automation](https://www.youtube.com/watch?v=Fqmw0G5hSl8\&utm_source=learning.contextqa.com\&utm_medium=docs_referral\&utm_campaign=contextqa_product_education\&utm_content=ship_integrations_mcp_claude). Exact clients, tools, and permissions depend on the authenticated workspace and connected server manifest.
{% endhint %}
Start with one bounded provider, project, repository, channel, or test target and expand only after verification.
## Choose the connection by outcome
| Desired outcome | Connection | Verify before relying on it |
| ------------------------------ | -------------------------------------------------------- | -------------------------------------------------------------------------- |
| Analyze code or a pull request | GitHub or GitLab | Repository scope, default branch, read access, and callback status |
| Create or link an issue | Jira, Linear, Azure DevOps, or another supported tracker | Project or team, issue type, required fields, and write boundary |
| Notify a team | Slack or another supported channel | Destination, message content, and evidence redaction |
| Gate a delivery | Test plan CI/CD configuration | Secret reference, environment, trigger, polling, timeout, and failure path |
| Reach a private application | Tunnel or remote browser | Named machine, route, browser support, health, and alert path |
| Work from an AI client | ContextQA MCP | OAuth identity, workspace, available tools, and read-before-write behavior |
## Connect an AI client with MCP
Open **Settings → MCP**, then choose ChatGPT, Claude, Claude Code, Cursor, Codex, or the generic method shown in your workspace. Copy the displayed configuration, complete OAuth in the client, and begin with a read-only request that confirms the organization and available tools.
```mermaid
sequenceDiagram
participant H as Human reviewer
participant A as AI client
participant M as ContextQA MCP
participant S as Ship workspace
H->>A: Ask for a bounded quality task
A->>M: Authenticate and inspect available tools
M->>S: Read approved workspace context
S-->>A: Return scoped records and evidence
A-->>H: Summarize facts, limitations, and proposed action
H->>A: Approve any change or execution
A->>M: Perform the approved operation
M-->>H: Return the result reference and evidence
```
{% hint style="warning" %}
Do not put passwords, provider credentials, private route addresses, or copied authorization values in an AI prompt. Use OAuth, provider secret stores, ContextQA environments, or an approved vault.
{% endhint %}
## Generate a CI/CD starting point
Open a validated test plan and select **CI / CD**. Choose GitHub Actions, GitLab CI, Bitbucket Pipelines, Azure Pipelines, CircleCI, or Jenkins, then select the intended environment and build source. Review the generated configuration before committing it.
Keep credentials in the pipeline provider's protected secret store. Prove both a passing and deliberately failing path before making the job a required check.
## Connect a private target
Use **Tunnels** when the target is not reachable from the public test infrastructure. Enroll the named machine, confirm its health and routes, optionally enable its browser host, then select that route from the test configuration. Configure a controlled alert destination so a down or degraded host is visible before a release run.
## Frequently asked questions
### Does MCP give an AI agent unrestricted access to our repositories?
No. MCP access follows the authenticated ContextQA workspace and tool permissions. Repository authorization remains a separate provider decision. The calling client can also have its own local permissions, which should be reviewed independently.
### Should every integration be connected during a trial?
No. One representative repository, tracker project, notification channel, and environment usually provide a cleaner evaluation and a smaller security boundary.
See [ContextQA MCP setup](/integrations-and-developer-tools/overview/installation-and-setup.md), [Integrations](/integrations-and-developer-tools/integrations.md), and [Operate Ship safely](/enterprise-application-testing/ship/security-and-controls.md).
## Connect the tools to a business outcome
Slack starts a contextual report; Linear, Jira, or ClickUp preserves the work item and configured trigger; GitHub or GitLab provides change context; MCP gives a compatible coding assistant access to supported test operations; CI applies the agreed release gate. Each connection has its own scope and authentication. Follow the [end-to-end practical guide](/enterprise-application-testing/ship/slack-linear-to-release.md) to verify the handoffs with a real report and candidate build.
---
# 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://learning.contextqa.com/enterprise-application-testing/ship/integrations-and-mcp.md?ask=&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.
