> For the complete documentation index, see [llms.txt](https://docs.currents.dev/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.currents.dev/ai/connect/ide-extension.md).

# IDE Extension

See CI runs and failures in VS Code or Cursor, and send a failure to your AI agent with its full context

The Currents extension shows your CI runs, failures and flaky tests in the editor, and sends a failure to your AI agent with its full context. It works in VS Code, Cursor and other editors that use the [Open VSX](https://open-vsx.org/) registry.

You need:

* A Currents account in an organization, or an API key
* VS Code 1.85 or later, or Cursor

## Install the extension

<figure><img src="https://3745692499-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FqmFDEiUa9mr11LUlxDnt%2Fuploads%2FTJCywA9tFiM8Kge9RU6g%2FCleanShot%202026-05-18%20at%2010.26.20%402x.png?alt=media&amp;token=20da0f47-737d-4ef8-a4eb-5b59e41954c6" alt="The Currents extension page in the VS Code Marketplace"><figcaption></figcaption></figure>

* [VS Code Marketplace](https://marketplace.visualstudio.com/items?itemName=currents.currents)
* [Open VSX Registry](https://open-vsx.org/extension/currents/currents) (Cursor and other VS Code forks)

Or search "Currents" in the Extensions panel of your editor.

## Sign in and pick a project

1. Open the Currents panel in the Activity Bar.
2. Click **Connect to Currents** and sign in from the browser. On a remote machine, run **Currents: Copy Sign-In Link** and open the link yourself.
3. Select a project. Its runs appear in the panel.

Where a browser sign-in is not practical, run **Currents: Set API Key** and paste an [API key](/dashboard/administration/api-keys.md) instead. If you keep both, the extension uses the sign-in.

<figure><img src="https://3745692499-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FqmFDEiUa9mr11LUlxDnt%2Fuploads%2F4P70XWxxBKlWu2zonpcr%2FCleanShot%202026-05-18%20at%2010.56.46%402x.png?alt=media&amp;token=3526c687-3402-43ce-8ade-2eb5a0cfd77f" alt="Currents extension settings with the Set API Key button"><figcaption></figcaption></figure>

## See CI failures

{% embed url="<https://player.mux.com/F8NYHWt01dG5KB00UGv7HOYNwzlKEIGBKqcLivLGhi3YA?metadata-video-title=Currents+IDE+Extension&video-title=Currents+IDE+Extension>" %}

### Run Feed

<figure><img src="https://3745692499-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FqmFDEiUa9mr11LUlxDnt%2Fuploads%2FUevYoPeLc14vijFgKe9n%2FCleanShot%202026-05-18%20at%2011.29.56%402x.png?alt=media&amp;token=417f11de-a5e8-479e-8908-050d1279e814" alt="The run feed in the Currents panel"><figcaption></figcaption></figure>

The latest CI runs, refreshed every 30 seconds.

* Filter by branch, author, tags, or status
* Toggle auto-refresh on or off
* Load older runs on demand

### Pull Requests

<figure><img src="https://3745692499-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FqmFDEiUa9mr11LUlxDnt%2Fuploads%2FpPSql7JWBDd6M2jgZrrt%2FCleanShot%202026-05-18%20at%2011.31.21%402x.png?alt=media&amp;token=6aee8ef0-20f3-4798-b927-3a40bf21cf9f" alt="Pull requests with their commits and builds"><figcaption></figcaption></figure>

CI failures grouped by pull request, with a timeline of commits and builds for each one.

* Filter by branch, author, tags, or status
* Toggle auto-refresh on or off

### Run Details

<figure><img src="https://3745692499-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FqmFDEiUa9mr11LUlxDnt%2Fuploads%2F7wa2bc3zPcaacYuJBzgK%2FCleanShot%202026-05-18%20at%2011.32.36%402x.png?alt=media&amp;token=e8b1f467-f3a0-4918-85e2-f3cd6b1de6b9" alt="Run details with a failed test, its error, and the Fix with agent and Trace Viewer actions"><figcaption></figcaption></figure>

Every spec and test of a run, with errors and stack traces for failed tests.

* **Fix All Issues** and **Fix with agent**: see [Fix with AI](#fix-with-ai)
* **Go to file**: jump to the test file and line in one click
* **Trace Viewer**: open the Playwright Trace Viewer inline (see [Inline Trace Viewer](#inline-trace-viewer) below)
* **Open in Dashboard**: open the full Currents dashboard for deeper analysis of the run, spec, or test attempt

## Fix failures with your agent

### Fix with AI

{% embed url="<https://player.mux.com/h7mDts1TwwARJseGM01u0001W638rJB9TKLlr01rkLRNtMI?metadata-video-title=Currents+IDE+Error+Context&video-title=Currents+IDE+Error+Context>" %}

The Fix actions send your editor's AI agent a prompt with the [failure context](/ai/agentic-testing/fix-failing-tests.md#what-the-context-contains): the error and stack, the test's history, and the error-context snapshot: the page's accessibility tree at the moment of failure.

| Action                                      | Sends                                                                      |
| ------------------------------------------- | -------------------------------------------------------------------------- |
| **Fix with agent** on a test                | That test's context                                                        |
| **Fix all failures in spec** on a spec file | The failed tests of that spec file, or its flaky tests                     |
| **Fix All Issues** on a run                 | The failures of the whole run                                              |
| **Fix With Agent** in Test Explorer         | The flakiness rate or average duration, for the agent to look into via MCP |

In an editor without a built-in chat, the prompt is copied to the clipboard.

### Analyze with Currents

**Analyze with Currents** appears above every `test`, `it`, `describe` and `context` in your code. It asks your agent to check the test's recent results in Currents for failures, flakiness and slowdowns, and to plan a fix. It is also a Quick Fix (Cmd+. or Ctrl+.) on any test line.

### The Currents MCP server

The extension adds the [Currents MCP server](/ai/connect/remote.md) to the editor, connected as you, so the agent can ask Currents for more after the first prompt. Signing out removes it. Turn this off with `currents.registerMcpServer` if you set up the MCP server yourself.

## View a trace

### Inline Trace Viewer

<figure><img src="https://3745692499-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FqmFDEiUa9mr11LUlxDnt%2Fuploads%2FSjNHI8znQkxxhccgSbfN%2FCleanShot%202026-05-18%20at%2011.36.23.gif?alt=media&amp;token=6d30e52a-7e6f-4540-b5f8-e1f5c25143a7" alt="The Playwright Trace Viewer in an editor tab next to the test code"><figcaption></figcaption></figure>

Click **Trace Viewer** on a failed attempt to open the Playwright Trace Viewer in an editor tab, next to your code. The trace streams from Currents, so there is nothing to download. Another trace opens in the same tab. **Open in browser** opens it in a browser window.

## Explore flaky and slow tests

<figure><img src="https://3745692499-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FqmFDEiUa9mr11LUlxDnt%2Fuploads%2F8Q8LYJyra0664dQ5TxQ4%2FCleanShot%202026-05-18%20at%2011.34.05%402x.png?alt=media&amp;token=12507b4a-a0fb-4537-af5c-77d1552fdbe9" alt="Test Explorer sorted by flakiness"><figcaption></figcaption></figure>

**Test Explorer** lists the flakiest and slowest tests of the project over 14, 30, 60 or 90 days.

* Sort by flakiness rate or average duration
* Jump to the test file and line in one click
* Send flaky or slow tests to AI for analysis and automated fixes
* Open in Dashboard for historical trends

## Settings

| Setting                          | Default                       | Description                                                                                 |
| -------------------------------- | ----------------------------- | ------------------------------------------------------------------------------------------- |
| `currents.filterByCurrentBranch` | `false`                       | Auto-filter runs by the current git branch when a project is selected                       |
| `currents.notifyOnRunComplete`   | `false`                       | Show an OS notification when a run finishes                                                 |
| `currents.analyzeTestDisplay`    | `always`                      | Show "Analyze with Currents" above each test (`always`) or only as a Quick Fix (`quickfix`) |
| `currents.apiBaseUrl`            | `https://api.currents.dev/v1` | API base URL (for self-hosted or enterprise setups)                                         |
| `currents.registerMcpServer`     | `true`                        | Auto-register the Currents MCP server for AI agents                                         |


---

# 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.currents.dev/ai/connect/ide-extension.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.