Download
RU EN DE FR ES PT 中文

Git Analytics — User Guide | YoBench

How to use the Git Analytics module in YoBench: connect GitLab cloud/self-hosted, pipeline & MR metrics, team statistics, AI code reviews.

What the Git Analytics module does

The module collects and visualizes analytics for your GitLab projects — pipelines, merge requests, commits and team metrics — without having to open the GitLab web UI. All data syncs to a local database on a schedule, so you can keep working offline and see a unified view of multiple projects/teams in one screen.

GitLab only (cloud gitlab.com and self-hosted). GitHub is reserved as a provider type in code but the adapter is not implemented yet.

What you get:

  • Unified CI/CD view — every pipeline of the selected projects in one table with filters by status and search by branch.
  • Code-review metrics — Lead Time, Cycle Time and average MR lifetime; pipeline success rate.
  • Per-developer stats — commits, insertions/deletions, MR count and comments per contributor + derived metrics (velocity, impact, collaboration, delivery rate, cycle efficiency, work-life balance).
  • AI review of new MRs — configurable AI reviewers with their own prompt and AI provider run automatically or on demand and persist results next to the MR.
  • Local cache — data lives in YoBench's encrypted database, available for filtering and analysis without network.

Project connection parameters

When adding a new project you specify:

  • Provider type — gitlab (currently the only working option).
  • Display name — how the project shows up in filters and tables.
  • Base URL — API URL: https://gitlab.com for cloud or your self-hosted URL.
  • Project path — e.g. group/subgroup/project.
  • Private token — token for the GitLab API. Needs read access to the repository, MRs and pipelines.
  • Git URL — clone URL (HTTPS or SSH); used for partial local processing (e.g. reading diffs).
  • SSH key (optional) — required when the Git URL is SSH.
  • Enabled — activity flag for sync.
  • Auto-review new MRs — when on, newly arrived MRs trigger the assigned AI reviewers automatically.

Global module settings

Configured under Settings → Git Analytics:

  • Enable Git analytics (gitAnalyticsEnabled) — master scheduler switch. Off by default — until enabled, projects are not synced automatically.
  • Sync interval (minutes) (gitAnalyticsPollMinutes) — how often the scheduler polls the API. Default 60 minutes.

Everything else — tokens, AI reviewers, auto-review — is configured per project or per reviewer.

AI reviewers

AI review is built around a "reviewer" entity. Each reviewer is:

  • Name — shown in the Review dropdown next to an MR.
  • AI provider — which provider from the AI Chat registry to use.
  • Prompt — instructions on how to analyze the diff (review style, focus areas, output format).

Reviewers are attached to projects in settings. With Auto-review on, the first time a new MR is synced:

  1. The module fetches the diff from the GitLab API.
  2. MR metadata + diff is sent to the assigned reviewers in turn.
  3. Each review result is saved to the database and visible in the Review panel next to the MR.

You can also trigger a review manually for any MR — the Review button opens a panel with reviewer selection and result.

UI capabilities

The main screen has four Analytics sub-tabs.

Pipelines

Table of pipelines for selected projects over the chosen period.

Columns: ID, branch, status (success / failed / running / canceled), source (push / merge / api), created at, duration.

Top summary: total pipelines, success, failed, success rate, average duration.

Filters: by status, search by branch. Clicking a row opens a panel with jobs and trace.

Merge Requests

Table of MRs with lifecycle metrics.

Columns: #, title, author, branch, state, created date, merged date, Lead Time, Cycle Time, comment count.

Summary: created MRs, merged, average lifetime, merge rate.

Filters: by state (opened / merged / closed), search by title or branch. The Review button runs or shows the AI review for a specific MR.

Project Stats

Metrics for the selected project over the period: commits, insertions/deletions, MR count, pipelines, branches, contributors. Day-by-day activity chart.

People Stats

Per-contributor table: commits, insertions, deletions, MRs created, MRs merged, comments. Sortable by any column. Extended metrics: velocity, impact, collaboration, delivery rate, cycle efficiency, work-life balance.

Period and project filters

Toolbar at the top of the page:

  • Period — week / month / quarter / year / all time. Default: month.
  • Project selector — one or many; empty means "all enabled projects".

Workflow

1. Add a project

  1. Open the Git Analytics module from the left sidebar.
  2. Switch to the Settings tab.
  3. Click Add project.
  4. Fill in Base URL, Path, Token, Display Name, Git URL.
  5. Optionally enable Auto-review new MRs.
  6. Save.

2. Turn on automatic sync

Under Settings → Git Analytics flip the master switch and set the interval in minutes. Without it, projects only refresh when you click Sync.

3. Sync data

A Sync button at the top of the page polls every enabled project: pulls pipelines, MRs, commits and branches. Progress is visible in Logs.

4. Explore analytics

Switch between Pipelines / Merge Requests / Project Stats / People Stats sub-tabs. Apply period and project filters.

5. Set up AI reviews

  1. Create an AI reviewer: name, AI provider, prompt.
  2. Attach the reviewer to a project.
  3. Turn on Auto-review new MRs — future MRs are reviewed automatically.
  4. For existing MRs, use the Review button in the list.

Next steps

  • Configure AI providers — without them AI reviews do nothing.
  • Use the Tracker to manage tasks alongside code analytics.
  • Larger teams can layer Health Check for separate uptime monitoring of GitLab instances.

Help and feedback

Found a bug or want to request a GitHub/Bitbucket adapter? Contact us via the feedback form.