Navigara
All Organizations

Cloudflare — Engineering Performance

Avg. perf / dev / mo (ETV)

+33.5%

0.76 → 1.01

Active contributors

+9.8%

133.0 → 146.0

Growth

+5.2pp

30.5% → 35.7%

Fixes

−13.0pp

30.2% → 17.2%

Cloudflare vs. 500 OSS Performance Index

Per-engineer ETV for Cloudflare plotted against the pooled 500 OSS Performance Index. Both series are 90-day trailing rolling averages scaled to a 30-day month, so the curves sit on the same scale (ETV / dev / mo) and can be compared point-for-point. Latest reading: Cloudflare is 48% below the index (1.01 vs 1.94 ETV/dev/mo). Baseline gap was 46% below.

Monthly reports

  • Highlights

    • Introduced **retained streaming agent tools** ([1b65ff55 · Sunil Pai]), enabling parent agents to orchestrate and inspect chat-capable sub-agents, significantly enhancing *agent orchestration* capabilities.
    • Implemented **alarm-backed APIs for sub-agents** ([8de0ce39 · Sunil Pai]), allowing for durable execution and recovery of long-running operations within the *agents* framework.
    • Enhanced *Miniflare* with **V2 protocol support for module fallback service** ([c07d0cb4 · James Opstad]) and the ability to gracefully handle missing static asset directories ([e5390082 · Dario Piotrowicz]), improving local development experience.
    • Delivered an **experimental programmatic API for Wrangler type generation** ([9a1f0149 · Ben]), providing flexible automation for *developer tooling*.
    • Introduced new *AI capabilities* including **`Think` agent `sendReasoning` controls** ([58ca2fc1 · Sunil Pai]) and **per-turn LLM observability telemetry** ([0ed42a90 · Sunil Pai]), offering finer-grained control and insights into agent behavior.
    • Improved *Pyodide* integration by enabling **processing of `.pth` files** ([628b230e · Hood Chatham]) and exposing the `send_email` API in *Miniflare's platform proxy* ([e653edf7 · Edmund Hung]).
    • Introduced **inline sub-agent event streaming** within the `agents-as-tools` example ([03620a67 · Sunil Pai]), enabling live progress and multi-turn interactions for helper sub-agents.

    Observations

    • *Maintenance* activity saw a significant increase, with the maintenance score rising 42% compared to the 5-month average (current: 43, average: 30). This was driven by extensive dependency updates, build system improvements, and a strong focus on documentation.
    • The *Grow* score decreased by 25% (current: 24, average: 32) compared to the 5-month average, indicating a slight shift from new feature development towards stabilization and refinement.
    • The *Waste* score decreased by 22% (current: 16, average: 20) compared to the 5-month average, suggesting improved code quality and reduced rework, despite several critical bug fixes.
    • A consistent pattern of **documentation enhancement** was observed, with multiple commits dedicated to improving clarity and completeness across various components, including *Cloudflare China Network* ([b21f6c36 · Pedro Sousa]), *Agents SDK* ([96e76737 · Sunil Pai]), *Wrangler API* ([ce079566 · Ben]), and *WAF rules* ([3b1d586e · Pedro Sousa], [0bf9922b · Pedro Sousa]).
    • Several **critical bug fixes** were implemented, addressing issues such as sub-agent bootstrapping failures ([6471cbd8 · Sunil Pai]), flaky React and chat concurrency tests ([2a151fd9 · Sunil Pai]), `fetch()` errors in *Pyodide* ([1249bd66 · Hood Chatham]), and regressions in *AI chat agent* tool-call replays ([8fb7c032 · Sunil Pai]).
    • Significant effort was directed towards **infrastructure and dependency management**, including upgrading Node.js to version 22 in CI/CD pipelines ([1f094412 · Hood Chatham], [b5ac54ba · Somhairle MacLeòid]), standardizing Vitest configurations ([32cd5765 · Pete Bacon Darwin]), and tightening internal peer dependency floors ([ca510d4f · Sunil Pai]).
    • The `require-description-when-disabling` ESLint rule was enabled across `create-cloudflare` ([31b14ac6 · Dario Piotrowicz]), `workers-utils`, and `workflows-shared` ([35518264 · Dario Piotrowicz]), indicating a push for improved code quality and maintainability standards.

Repositories

Active repositories ranked by average performance per developer per month (over the last 90 days). The chart shows monthly performance composition — each repo as a stacked layer, with the top of the stack representing total org performance per month. Top 9 repos shown; the remainder is aggregated as “Other”.

Repository
agents61196.6
+1294%since Q2 2025
sandbox-sdk4302.5
+348%since Q2 2025
foundations141.3
+2313959%since Q2 2025
lol-html141.2
−5%since Q2 2025
workers-sdk381020.9
+11530%since Q2 2025
telescope250.8
+396%since Q4 2025
pingora6110.6
+109%since Q2 2025
containers230.5
−80%since Q2 2025
workerd42640.5
+16%since Q2 2025
workers-rs220.4
+47%since Q2 2025
cloudflare-docs101990.3
−5%since Q2 2025
cloudflared310.2
−9%since Q2 2025
api-schemas100.0
Company total13 repositories
146unique devs
445ETV total
1.01ETV / dev / mo
+134%since Q2 2025
Performance (ETV) is the sum of every repository above. Active devs at the company level counts unique contributors across all repos, so a contributor working in multiple repos is counted once here but appears in each repo's row (the per-repo column will sum higher). ETV / dev / mo = Company ETV ÷ unique devs ÷ 3 mo. The "Since start" column compares each repo's Q1 2026 quarterly performance to the first quarter it had any activity — for repos that existed in Q2 2025 (when this index began), that's Q2 2025; for younger repos it's the quarter they actually started. The company row uses Q2 2025 as the baseline since the index itself began then.

Performance Growth vs Active Contributors

Engineering performance is outpacing team growth by . Left axis shows total performance score, right axis shows active contributor count. The gap between curves represents productivity gains — more delivered per person, not just more people. Unit: Engineering Throughput Value (ETV).

Cost per Performance Unit

−54%

If performance per engineer more than doubled, each unit of engineering performance now costs approximately 54% less than at the baseline 90-day window (ending 2025-06-29). This is a directional estimate — the exact figure depends on fully-loaded engineer cost, but the direction is unambiguous.

Effective Capacity Added

+171 engineers

At today's productivity, the current 146-person team delivers the performance equivalent of 317 engineers at the baseline 90-day rolling window (ending 2025-06-29). That's roughly 171 engineers worth of capacity added through productivity gains, not hiring.

Performance Composition

Stacked bars show total complexity performance split into Growth (new value), Maintenance (sustaining systems), and Fixes (rework). The yellow line overlays performance per contributor — rising line means each engineer is delivering more, regardless of team size changes. Unit: Engineering Throughput Value (ETV).

CapEx vs OpEx

Monthly CapEx vs OpEx split. CapEx (capitalizable investment) is Growth — new features and capabilities. OpEx (operating expense) is Maintenance plus Fixes — keeping the lights on and reworking what's already shipped. The yellow line is the CapEx share, a quick read on how much of the month went into building new vs sustaining existing. Unit: Engineering Throughput Value (ETV).

Hours per Repository

Trailing 90-day window (64 working days). Org-level capacity is allocated to each repo by its share of org performance, then split CapEx / OpEx by that repo's own Growth vs Maintenance + Fixes mix.

api-schemas100.0%0.0%
telescope65.2%34.8%
pingora60.2%39.8%
containers59.4%40.6%
foundations48.0%52.0%
workers-rs46.9%53.1%
agents44.0%56.0%
workerd42.8%57.2%
sandbox-sdk41.8%58.2%
workers-sdk37.2%62.8%
cloudflared19.2%80.8%
cloudflare-docs13.5%86.5%
lol-html5.1%94.9%
Total35.7%64.3%

Fix Burden Distribution

Monthly rework volume broken down by who did it. Top contributors carry their named slice; everyone else is rolled into Others. Use this to spot whether fix work is concentrated on a small group (bus-factor risk) or distributed across the team.

Fix authorship over time

Monthly fix activity in this scope, split by who fixed the bug. Darker emerald = the original author fixed their own bug; lighter emerald = someone else cleaned it up. A persistently high 'fixed by another author' share is a signal of bug-debt landing on the team rather than its author. X-axis is fix-time — bugs introduced but not yet detected don't appear.

Self-fix share
26%
Total bug impact
346.4

Quarterly Summary

The raw numbers behind the charts: commits analyzed, active contributors, total performance, performance per developer, and the Growth / Maintenance / Fixes split for each quarter.

Quarter
Q2'252,138126179.60.528.4%59.5%12.1%
Q3'252,754137256.520.630.2%51.5%18.4%+43%
Q4'252,410130253.180.627.4%51.8%20.8%−1%
Q1'262,668139419.64136.4%37.6%26%+66%

Top Contributors

Contributors ranked by performance per month (Growth + Maintenance + Fixes), over the last 90 days normalized to a 30-day calendar month.

The best way to measure AI efficiency

Sample

A preview of the Navigara engine running on a sample organization. The numbers below are illustrative, not part of the OSS500 benchmark above.

Measure

Score every commit by depth

GitHub commits are weighted by what it took to write them, not by lines of code. The result is ETV per developer per month.

SourceGitHub

Spend

Tie ETV to cost

AI token bills and seat costs are pulled per team and divided by the ETV produced. The result is your true cost per unit of work.

SourceToken usage + finance

Map

Tie work to objectives

Each ETV is mapped to your Jira epics and labels, so you can see what's key-aligned, aligned, or unmapped capacity.

SourceJira

Performance

ETV delivered per developer / month

9.4ETV / dev / month

5.6 below target
target 15
Jun '25Dec '25May '26
  • Non-AI 5.8
  • AI 3.6

AI Efficiency

AI spend per ETV unit delivered

$4.20

$0.60 over target
target $3.60
Jun '25Dec '25May '26
  • Cost / ETV $4.20

Objective Alignment

Share of work mapped to key objectives

51%

24 pts below target
target 75%
Jun '25Dec '25May '26
  • Key-aligned 28%
  • Aligned 23%