---
title: "A better GitHub run comment"
description: "The Currents comment on your pull request now updates while the run is in progress, shows each failed test with its history and trace, and gives you and your agent a prompt to fix it."
canonical: "https://currents.dev/changelog/github-run-comment"
updated: "2026-10-19"
---

# A better GitHub run comment

2026-10-19, Improved (tracks: platform)

The Currents comment on your pull request now updates while the run is in progress, shows each failed test with its history and trace, and gives you and your agent a prompt to fix it.

---


Currents now posts its run comment when the run starts and edits it as the run continues, and it can add inline comments on the failing lines of changed spec files. The header shows the result, the failed and flaky counts, the test and group counts, the Playwright version, the duration and the commit. The table lists up to 10 groups.

![Currents run comment on a pull request: the run's failed and flaky counts, a table of the groups with their test counts, and the first failed test with its error, last 30 days stats, frames of the failed attempt and links to the trace, page snapshot, screenshot, video and context for agents](/img/changelog/github-run-comment-1.png "The run comment after a failed run, with the first failed test")

## Failed and flaky tests

The comment lists up to 10 failed and up to 5 flaky tests, in a collapsed list. Each shows its spec file, its groups, the duration of its failed attempt and, when retried, which attempt failed, such as `failed attempt 2 of 2`.

A failed test shows an error summary: the message's first line and up to five lines after it. A different error in other groups gets its own line, with a **View** link.

The history line compares with the pull request's base branch:

`Last 30 days: failed 18 of 85 · flaky 0% · on base/main: failed 0 of 15`

When the run has failed tests and Fix links are on, failed and flaky tests get links: Trace, Page snapshot, Screenshot, Video and Context for agents, which also links to Playwright's error context. For a Playwright test whose trace has screenshots, the links start with a row of frames from the trace up to the failure, with the failure frame marked in red.

## Fix with AI

When the run has failed tests, including canceled and timed-out runs, the comment has a **Fix with AI** section:

- **For you** — a Fix link that opens the failure in Claude Code, Cursor, Copilot and 4 more
- **For your agent** — a one-line prompt in a code block with a copy button: `Fix the failed tests in <Fix link>.md`
- **Full prompt with failure context** — collapsed, and left out when the comment would pass GitHub's size limit

The last line reads "Fix links are public · expire &lt;date&gt;, &lt;time&gt; UTC". With Fix links off, the section says "Fix links are off" and links to the run in Currents, and the prompt gives the run ID for an agent connected to the Currents MCP server.

![Fix with AI section of a run comment: a link that opens the failure in Claude Code, Cursor, Copilot and other agents, a one-line prompt for an agent, the collapsed full prompt and the date the Fix links expire](/img/changelog/github-run-comment-2.png "Fix with AI with Fix links on")

Anyone with a Fix link opens it without signing in. In a public repository, anyone who can see the pull request can open the links, and the comments are public too. To keep the links out of a public repository's pull requests, turn off **Add Fix Links to Comments** in that project's GitHub App settings.

## Inline comments on spec files

With **Post Error Details on Spec Files** on, a failed test gets a comment on the spec file line where its error points, when that line is in the diff. An error in another file, such as a page object, gets none. The comment shows the groups where the test failed, one error summary per different error, and a Fix with AI section.

Currents covers up to 10 changed spec files, most failures first, and the first 5 failed tests in each. Later runs update the comment: it lists up to 5 **Previous failed runs** and moves when the error moves. A finished run that passes the test deletes it. Currents does not resolve review threads.

![Currents review comment on a changed line of a spec file: the test failed in mobile, desktop and tablet, with one error for mobile and tablet and another for desktop, and a Fix with AI section](/img/changelog/github-run-comment-3.png "An inline comment on the failing line, with a different error in two groups")

## New settings

The settings apply to each Currents project, which reports to one repository:

- **Which Runs Get a Comment** — **Always post** (the default), **Only when there are failures** or **Only when there are failed or flaky tests**
- **Existing Comment** — **Update it** (the default): a newer run edits the existing comment, so each project and framework keeps one comment for the latest run that started. **Post another**: each run posts its own comment and keeps the earlier ones.
- **Post Error Details on Spec Files** — adds the inline comments, off by default
- **Add Fix Links to Comments** — adds Fix links to the comments, on by default
- **Fix Link Expiry** — 1, 3 or 7 days, 1 day by default. A change does not affect links already posted.

With **Update it**, a run that **Which Runs Get a Comment** leaves out can still update an earlier run's comment. With **Only when there are failures**, a passing run after a failed one updates the failed run's comment to show the pass.

![GitHub App settings in Currents: the GitHub organization and repository, then the Commit Status, Run Comment, Inline Comments and Fix Links sections with their settings, and the Disconnect Repository and Save buttons](/img/changelog/github-run-comment-4.png "The GitHub App settings after connecting a repository")

Learn more:

- [GitHub App](https://docs.currents.dev/resources/integrations/github/github-app)
- [Inline comments on spec files](https://docs.currents.dev/resources/integrations/github/github-app#inline-comments-on-spec-files)
- [Shareable AI context for test failures](/changelog/shareable-ai-context)

