the open-source toolkit for product managers
Sharper thinking.
Not just faster output.
mogkit installs PM craft into your terminal — a library of methodology-backed skills that run in Claude Code, a thin CLI that scaffolds your workspace, and a growing library of AI-for-PM workflows you can ship today.
It won't write your PRD. It'll make you write a better one.
$ npx mogkit init my-workspace
Craft, not output
No skill writes the finished deliverable.
Every skill produces a reasoning scaffold: what your evidence supports, what you're assuming without proof, what to validate next. You write the doc — and the doc is sharper. We sell judgment, not typing.
Local and yours
Runs in your terminal. MIT-licensed.
Version-controlled workspace. Your research is an asset you own and grow, not data in someone else's cloud. No SaaS, no login, no lock-in. The CLI never calls a model.
The on-ramp
From "lives in Notion" to "works like an engineer."
Every step is small. Every step earns the next. We don't gatekeep the terminal — we make it approachable. Setup 101 is genuinely followable by someone who has never opened one.
By the craft, for the craft
Methodology-backed. Community-built.
Continuous discovery, Working Backwards, the real traditions. Skills are plain markdown; workflow entries are MDX; both follow documented specs. The library compounds as PMs contribute to it.
One deep system, many sharp tools
The five wedges
Discovery
Customer research as a versioned, compounding asset. mogkit's deepest system lives here.
Strategy
Six-pagers, memos, and metric trees that survive contact with a smart skeptic.
Planning & Roadmapping
Specs that hold up under attack and roadmaps that don't lie about the tradeoffs.
GTM & Launch
Launches fail for boring reasons. The fix is also boring: a pre-flight interrogation specific to this launch, not a generic checklist.
Stakeholder & Conflict
A stalled decision is rarely a logic problem. It's an incentive problem you haven't named.
Recipes + mechanism
Workflow Library
Implementable AI-for-PM workflows. Each entry is both a copy-paste recipe and a "how it works" explainer that teaches the mechanism. We never ship a black box.
Weekly Discovery graph diff: surface what changed in your research
A weekly cron compares this week's graph to last week's and surfaces new Pains, new Segments, new Assumptions, and resolved gaps — the diff a PM should actually read.
Support → docs agent: Intercom conversations → Mintlify PRs
A scheduled agent reads recent Intercom conversations and opens a docs PR with proposed edits + a gap report. A human reviews and merges.
Automated QA gate for Linear issues
Every Linear issue gets an automated QA pass against its own acceptance criteria before review.
Get started
Two commands and you're set up.
If you've used a terminal before, two commands and you're set up. If you haven't, Setup 101 walks you through every step from opening the terminal to running your first skill.
$ npx mogkit init my-workspace
$ cd my-workspace
$ open this folder in Claude Code
$ run your first skill