Runs on top of GitHub · AI-native via MCP

Project management, native to GitHub.

Plan sprints, track pull requests and issues, log your working time, and let AI agents run the busywork over MCP — all synced two-way with your repos, so the board never drifts from the code.

Start free See how it works

Free to start · Connect a GitHub repo in minutes · No credit card.

acme / web-app · Sprint 14Labels synced

Ready

#482 Fix billing webhook

ready

In progress

#479 Public doc sharing

00:42:18

#477 Sprint reports

testing

Passed

#471 OAuth refresh

passed

MCP agent generated 4 test cases for #479 and logged 12m.

Plan, track, time, and automate delivery — in one place, on top of GitHub.

NORTHWINDQuantaBYTEFORGEHelixOAKLINE

One workspace, three superpowers

GitHub is where the code lives. KazTrack is where the work gets planned, tracked, timed, and automated around it.

PM built on GitHub

Sprints, pull requests, issues and milestones, scoped per project. Workflow states map to GitHub labels and sync back automatically — plan where the code already lives.

Track working time

Start/stop timers on the work itself. See exactly where hours go with reports by person, day, repository, pull request, and label. Turn "are we on track?" into a number.

AI-native via MCP

A built-in MCP server exposes the whole workspace to Claude Code, Codex, or any MCP client. Agents read PR diffs, run test cases, update sprints, edit docs — with the user's exact permissions.

From repo to report in three steps

And an MCP agent can assist at every step.

Connect GitHub

Plan & track the work

Build sprints, claim and assign PRs, move them through states, write test cases, and collaborate in real-time docs.

Time it, ship it, report it

Log working time, trigger Forge / xCloud deploys per PR, and pull reports by tester, day, repo, PR, or label.

Everything your delivery needs

A complete toolkit around your GitHub workflow.

Pull-request workflow

Claim PRs, assign owners, advance states (ready → in progress → passed/failed), and sync them as GitHub labels.

Sprints, milestones & issues

A planner tied back to repos and PRs. Create, verify and publish issues — keep scope and progress in one view.

QA suite

Test cases & suites, assign-to-PR, record results — with optional AI-assisted generation and execution.

Real-time docs

Runbooks and specs with multiplayer editing, comments, mentions, version history — and public read-only share links.

One-click deploys

Trigger and watch Laravel Forge and xCloud deployments per pull request, without leaving the workspace.

Granular RBAC

Owner / manager / tester / viewer plus custom per-project roles. Permission-driven access, mirrored to MCP tokens.

GitHub-native

Your board never drifts from your repo

Move a pull request from ready to passed in KazTrack and the matching GitHub label updates instantly. Comments post to the real PR. Stop reconciling two sources of truth — there's only one.

Two-way label sync between workflow states and GitHub
Per-project repositories, members, and roles
Sign in with GitHub or a project bot token

#479 Public doc sharingin progress

GitHub · label in-progress applied

This week37h 12m

web-app21h

billing-svc9h 40m

docs6h 32m

Time tracking

Know exactly where the hours go

Timers live on the work — start one against a PR or task, stop it when you context-switch. Roll it all up into reports by person, day, repository, pull request, and label, and answer "are we on track?" with evidence.

Start/stop timers tied to real work
Reports by tester, day, repo, PR, label
Exportable summaries for stand-ups and billing

AI-native · MCP

Let agents do the busywork

KazTrack ships a built-in MCP server, so Claude Code, Codex, or any MCP client can operate the whole workspace — read a PR's diff, generate and run test cases, update a sprint, edit a doc, kick off a deploy — always bounded by the token owner's exact permissions.