Plan it. Track it. Test it.
Honest Cheetah sits on top of GitHub Projects and does the three things a board can't do by itself: help you plan and manage the work, tell you how it's really going, and manage your QA testing — cases, runs, and results, right next to the work. Your data stays in GitHub. HC reads it, and writes only what you confirm.
Sign in with GitHub Explore the demo orgPlan
A backlog with opinions. Requirements, tasks, and sub-issues you can create, reorder, and edit without leaving HC. A one-click refinement flow — Proposed → Not Ready → Ready — installed in front of Todo on a new project or one you already have, so intake is trackable from day one. No sprint jargon required. Why refinement? →
Project factory
A private repo, a linked Projects board, refinement statuses, iterations, and metric settings — one page, two minutes. Sample projects across five domains if you want to kick the tires first.
Requirements and tasks
Create requirements singly or in batches. Break them into tasks with remaining-hours tracking. Sub-issue progress rolls up as counts — never percentages.
AI that helps you write better stories
Score a story for verifiability (can a tester tell what passing means?), get a rewrite that fixes the findings, draft a requirement from messy notes — or have it propose backlog items from the project itself. Every suggestion is a draft. You confirm every write.
Track
How it's really going — from your team's own history, not anyone's estimates.
Flow metrics
Cycle time, lead time, throughput. You tell HC which statuses mean "started" and "finished"; it computes the rest from your status history.
"When will it be done?"
Monte Carlo forecasting over your throughput. When will item #14 finish? How much fits by the date? A range with a confidence level — the honest answer.
Explained in plain English
AI narrates what the numbers mean. It never does the arithmetic — the metrics are deterministic; the explanation is the only part the model touches.
Test
GitHub famously has no test management. This is it. Test cases linked to the requirements they verify, organized into suites, recorded as runs and results — living next to the work instead of in a separate tool nobody opens.
Test cases, where the work is
Create, edit, and organize test cases against requirements. Suites for what gets run together. A breadcrumb from every case back to the requirement it verifies — so coverage is a thing you can see, not a thing you assert.
Runs and results
Record test runs and results as they happen. Optionally pick the GitHub Actions run — and the environment, if you use them — you tested against, so the record says exactly what you tested, right down to the SHA. Completed runs stay put; corrections are append-only.
AI drafts the test cases
Point it at a requirement and get candidate test cases — skipping the ones you already have. Drafts you edit and confirm. Nothing reaches GitHub without a click.
Every result also records who ran it and as what — human, agent, or automation — so when it matters, you can tell the difference between "verified" and "vibes." The chain of custody, if you want to go deep →
Work in GitHub. Work in HC. Same data.
Everything project management — requirements, tasks, sub-issues, statuses, priorities, iterations, backlog order — is GitHub Projects and Issues data, written back to GitHub with your own credentials. Open the same board in GitHub and it's all there. No shadow copy. No lock-in. (Well — apart from your data being locked into GitHub. But you picked GitHub. Thems is the breaks.)
What HC adds is purpose-built screens: a backlog that's actually a backlog, refinement you can see, task creation that does the whole job in one step. The rule is simple: if GitHub can store it, GitHub stores it.
The one exception, stated plainly: test cases, runs, and results live in Honest Cheetah — because GitHub doesn't have a QA testing product to put them in. That's the gap HC fills. Your test cases still link back to the GitHub issues they verify.
Look before you connect anything
The demo organization has four sample teams and twelve weeks of history — explorable without connecting a real GitHub org. Read-only, no strings. Ready to try it with your own org? The project factory will spin up a sample project — five domains, pick your tech stack — so you can see it working before you put real work in.
Coming soon
The alpha is a working product, not a preview. But here's what's on the bench, roughly in order:
- Plan a requirement — AI proposes a task breakdown and test cases for a requirement; you confirm and publish in one click.
- Definition of Done you can audit — define done-criteria per project; watch each requirement's criteria light up as tests pass and the branch merges.
- Story splitting — AI suggests the seams in a too-big requirement and proposes the thinner stories, in order.
- Impediments — track what's blocking work, and how long it's been blocking.
- Defect tracking — bugs recorded against the exact commit, build, and deployment where you found them, then routed into the backlog as work you can schedule.
- Sprint taskboard and iterations view — the sprint-shaped views for teams that work in sprints.
- Sprint forecast — your ordered backlog annotated with which iteration each item will probably land in.
- Burndown — arrives with the sync infrastructure, because burndown needs history that starts accruing the day you install the GitHub App.
- Test evidence capture — screenshots per step, straight from the browser, attached to the run.
- One-click status reports and sprint review decks — meeting-ready, generated from the board.
Tell us what's wrong with it
Alpha means we'd rather hear it now. What's confusing, what's missing, what you'd pay for, what you'd never use. feedback@honestcheetah.com
Free for teams of 5
Five seats free, no trial, no countdown. During the alpha, everything is free for everyone. Where pricing is headed →