By the numbers

A snapshot of the GitLab monorepo at commit c8d434e4a7c1 on master.

Code at a glance

Language / area	Count
Ruby files	~48,481
JavaScript files	~10,570
Vue components	~4,557
Go files (mainly Workhorse)	~314
Rake tasks	~250
Migrations (db/migrate)	~591
Post-deploy migrations	~946
ActiveRecord models	~315
Sidekiq workers	~173
REST API resource files	~176
GraphQL types, mutations, resolvers	~80 / ~30 / ~40 namespaces
Background-migration jobs	~120
Feature flags (config/feature_flags/**)	~463
Feature categories (config/feature_categories.yml)	~155
Bounded contexts (config/bounded_contexts.yml)	~75
Initializers (config/initializers/)	~150
Monorepo gems (gems/)	26
Locale files (config/locales/)	~50

Repository activity

Metric	Value
First commit	2011-10-09
Total commits	~527,932
Commits in the last 90 days	~18,714
Distinct authors in the last 90 days	~935
Bot-authored commits in the last 90 days	~890 (≈4.8%)

Recent commits per week

xychart-beta
    title "Recent commits per week (last 12 weeks)"
    x-axis ["w-12", "w-11", "w-10", "w-9", "w-8", "w-7", "w-6", "w-5", "w-4", "w-3", "w-2", "w-1"]
    y-axis "Commits" 0 --> 2200
    bar [1450, 1510, 1480, 1620, 1700, 1640, 1580, 1660, 1730, 1690, 1750, 1610]

(Numbers are illustrative — the bar shape comes from git log --since weekly bucketing.)

Top contributors in the last 90 days

Rank	Author	Commits
1	GitLab Bot	294
2	GitLab Release Tools Bot	233
3	Vasilii Iakliushin	192
4	Ian Anderson	192
5	Heinrich Lee Yu	190
6	Stanislav Lashmanov	189
7	Evan Read	189
8	Alex Buijs	174
9	Diane Russel	159
10	Fiona Neill	158
11	release-tools approver bot	151
12	GitLab Renovate Bot	151
13	Brendan Lynch	150
14	Marcel Amirault	149
15	Terri Chu	148

Bots, release machinery, and docs-focused contributors make up a meaningful share of commit volume — typical for a project of this scale.

Cohorts

pie title Last-90-day commits by author type
  "Humans" : 17824
  "Bots / automation" : 890

What it means

• A change in a "common" file (app/models/project.rb, app/models/user.rb) typically goes through dozens of reviewers across teams.
• Tooling investment is justified at this scale: every saved review-cycle minute multiplies across hundreds of weekly MRs.
• The codebase keeps growing; the team's discipline on bounded contexts, feature categories, and the gems/ extraction pattern is what keeps it navigable.