Ship better
AI code.

AI coding analytics for engineering teams: the agents, models and skills behind your merged pull requests.

Start freeGet started

Free for 3 seats · No credit cardStarter covers 3 seats · Pro is $49 a month for 5

Works with
  • Claude Code
  • Cursor
  • VS Code
  • Codex
  • Hermes
  • Grok Build
  • OpenCode
  • Freebuff
Sample data

USED BY ENGINEERING TEAMS AT

Flyps Charm Wildfires Geoforge MisaLabs

See which coding agents, models and skills your team uses in Claude Code, Codex, Cursor and Copilot, what they cost, and which sessions end in merged pull requests.

01

See your team’s AI stack.

Models, harnesses, skills, and agents. See the setup behind the work.

Sample workspaceClick a tool or a number to open its details.

Winding / Teams Last 30 days

Zest

3 Members 1 Agent
Team overview
67%
AI Adoption▲ 34 pts
240
PRs Merged▲ 101
96.1h
Median AI Time▲ 9.5h
57.2h
Build Hours▲ 20.3h
This week
Member AI PRs/wk AI Task Plan Impl Last Standup AI Stack
22.8/wk28 min ago
15.8/wk12 min ago
4/wkAug 14
—2 hours ago
Current AI models & harnesses As of Sep 8
Claude CodeFable 5.1 (max) (with fallback)70.4 · Coding Agent Index
CodexGPT-6 Astra (max)67.0 · Coding Agent Index
Muse CodeMuse Spark 1.3 (xhigh)64.2 · Coding Agent Index
Grok BuildGrok 4.5 (high)64.1 · Coding Agent Index
Kimi Code CLIKimi K362.6 · Coding Agent Index
OpencodeGemini 3.8 Flash (high)61.2 · Coding Agent Index
Cursor CLIComposer 2.5 Fast38.3 · Coding Agent Index

Demo setups use current model families. The pairings above are the combinations benchmarked by Artificial Analysis; rankings measure performance, not popularity.

Intelligence Index

Illustrative daily trends · Other totals: last 30 days. Build Hours: this week, model-estimated activity including overlap; not engineer hours.

02

Share the best AI skills.

Discover a teammate’s shortcut. Try it in your next session.

Sample workspaceThe email your team gets every Monday.

Weekly email

Your team’s weekly edge

Zestto your teamEvery Monday

AI stack worth sharing

  • Agree on the approach before coding.

    The brainstorming skill helps the team make a plan together.

    - [ ] **Plan** before coding.12 → 21 sessions · 5 → 8 teammates
  • Let agents work safely in parallel.

    using-git-worktrees gives each agent its own space to work.

    - [ ] One agent, one `worktree`.4 sessions · 2 → 3 teammates
  • Catch missing requirements early.

    A quick gap analysis finds what’s missing before anyone starts building.

    - [ ] **Find** missing cases.38 of 134 sessions · 9 teammates

03

Celebrate what ships.

Make the work visible. Share the practice that made it possible.

Sample workspacePick a teammate to see their week.

Winding / People Sep 4, 2026

Adrian Cardoso

Team ZestEng
Shipping momentum
32 AI tasks11h 23m AI time13 agents ran
32
AI Tasks▲ 8
11h 23m
AI Time▼ 1.2h
13
Agents Ran▲ 3
6.2h
Build Hours▲ 1.1h
3.8h
Bug Hours▼ 0.9h
1.4h
Process Hours▼ 0.4h
🧠 AI Standup Summary
Built with:Claude CodeClaude Opus 5
  • 🚀 Shipped deep-linkable session URLs.

    Added stable links and workspace-aware navigation for AI sessions.

  • 🔎 Investigated a production dispatch spike.

    Traced the timeout path and created a tracking task with the findings.

  • 🤖 Kept the review → fix → verify loop moving.

    A review loop worth sharing.

AI stack worth sharing
pr-watch28

A review loop worth sharing.

Adrian uses pr-watch to keep reviews moving. Turn that repeatable workflow into a team habit.

About this sample

Example standup · Illustrative daily trends. Kudos stay in this preview. *Model-estimated activity, including overlapping automation; not engineer working hours.

04

Fix agentic loop bottlenecks.

From CI to review to merge. See where the agentic loop loses time.

Sample workspaceHover a stage to see why it is slow and what to change.

Winding / Agentic loop Sample week

Delivery loop

Team Zest
41 merged PRs
Code→Check→Review→Merge→Live
Pipeline stageMedian
CI wall time 12.4mRuns per PR 2.5Test / verify activity 18h / week
Approve → merge28m per PR

The largest delay comes after approval: the work is ready, but the merge is still waiting.

Give the agent a clear merge rule: merge when approval and all required checks are green.
About this sample

Sample from Zest’s pipeline report · Stage medians overlap and do not add up. Test / verify activity is a weekly session-flow estimate, not pipeline wall time.

05

Increase throughput weekly.

A shared view of the week, with one clear next step for the team.

Sample workspacePick a week, or open the weekly email.

Winding / Team metrics Sep 4

Team weekly

Team Zest
vs prior week
13/14 AI Adoption55 PRs Merged10.3h Median AI Time
13/14
AI Adoption▲ 3
55
PRs Merged▼ 8
10.3h
Median AI Time▼ 3.5h
57.2h
Build Hours▲ 20.3h

Agents Ran

Planning43
Implementation49

AI Plan Stack

Top Dev46
p256
About this sample

Sample team week · Earlier trend points are illustrative. *Model-estimated activity, including overlapping automation; not engineer hours. Deltas use the email’s unrounded totals.

06

Increase AI experimentation.

New models, harnesses, and skills — see your team’s AI stack evolve.

Sample workspacePress Next update to watch the stack change.

Winding / Team experimentation Example week

Zest

3 Members 1 Agent

Your team’s AI stackNew tools appear here as teammates try them.

Member AI PRs/wk AI Task Plan Impl Last Standup AI Stack
22.8/wk28 min ago
15.8/wk12 min ago
4/wkAug 14
—2 hours ago
Current AI models & harnesses As of Sep 8
Claude CodeFable 5.1 (max) (with fallback)70.4 · Coding Agent Index
CodexGPT-6 Astra (max)67.0 · Coding Agent Index
Muse CodeMuse Spark 1.3 (xhigh)64.2 · Coding Agent Index
Grok BuildGrok 4.5 (high)64.1 · Coding Agent Index
Kimi Code CLIKimi K362.6 · Coding Agent Index
OpencodeGemini 3.8 Flash (high)61.2 · Coding Agent Index
Cursor CLIComposer 2.5 Fast38.3 · Coding Agent Index

Demo setups use current model families. The pairings above are the combinations benchmarked by Artificial Analysis; rankings measure performance, not popularity.

Intelligence Index

Illustrative team activity · New tools appear as teammates try them. Demo identities.

SIMPLE PLANS. ROOM TO GROW.

Start small.
Bring the whole team.

One workspace price. Seats included. Credits shared.

Starter

Try AI analytics

$0 per workspace per month

3 seats included

Get Started

No credit card needed.Switch to Pro from Billing when you need more.

  • Coding agent analytics
  • Models & Skills analytics
  • AI cheatcodes

Max

Scale AI analytics

$149 per workspace per month

10 seats included

Choose Max

Starts free on Starter, no card. Switch to Max from Billing.Every workspace starts on Starter. Switch to Max from Billing.

  • Agentic product management
  • AI Feature + Ticket analytics
  • Cloud agent + Hermes

Enterprise

Custom credits + history · SSO/SCIM + audit controls · Data governance + custom retention · SLA + dedicated success

Talk to us

A head start for startups.

50% off Pro or Max for 6 months. Pre-Series A, under $2M revenue in the past 12 months.

Apply for the discount

A FEW THINGS TO KNOW

Questions

What does Zest help my team do?+

Zest brings together AI coding activity, the tools and skills your team uses, and GitHub context. It helps you share useful practices and see what to improve next.

What are “build hours”?+

These are model-estimated activity hours, including overlapping automation. They are not employee working hours, timesheets, or a measure of a person’s value.

Can we start for free?+

Yes. Starter is $0 per workspace per month with 3 seats included. You can choose a paid plan as your team’s needs grow.

What metrics does Zest track?+

Per engineer and per team: AI tasks, AI time and agents run each day, the models, skills and MCP tools each session used, tokens and list-price cost, and pull requests opened and merged, with PR throughput and cycle time from GitHub. Zest has no DORA metrics.

Which tools can we connect?+

Claude Code, Cursor, VS Code (GitHub Copilot Chat), Codex, Hermes, Grok Build, OpenCode and Freebuff, each through its own Zest plugin or extension, and GitHub for pull requests. Each one has a setup guide in our docs.

What data leaves an engineer's machine?+

Each coding-agent session's prompts and replies, model, token counts, skills and tool calls, after secrets are redacted on the machine. Zest collects no code diffs. The workspace owner can stop collecting prompts, replies or GitHub events, and keep data for 12 hours, 1, 7, 30 or 90 days; 90 days is the most, and an engineer can only make their own settings stricter.

Which Git hosts does Zest work with?+

GitHub only, through the Zest GitHub App. It reads pull requests, commits and their metadata, and never writes to your code. Zest has no GitLab, Bitbucket or Azure DevOps integration.

Is our data used to train AI models?+

No. We do not sell customer data, and no customer data is used to train AI models. Our sub-processors page names each company that handles data for us and what it does.

How do we set Zest up?+

An admin installs the Zest GitHub App and picks the repositories. Each engineer then installs the Zest plugin or extension in their coding agent and logs in, and their sessions are recorded from then on.

Who makes Zest?+

Zest is made by a small team at Winding Labs. Ilya Volodarsky, who co-founded Segment, is one of its co-founders. Several products share the name; this one is the AI coding analytics product at meetzest.com.