Paul Ducey
DuceyWORK ← All Skills
Skills / kaizen-charter
✓ Free · MIT License Manufacturing Healthcare

Kaizen Charter

A kaizen event that fails usually fails before day 1: a problem nobody's watched yet, a team that can't get freed, a process too unstable to improve this week. This skill coaches the readiness questions first, then drafts a complete charter a sponsor could sign.

⬇ Download Skill Readiness questions ↓

Complete Charter: 11 Sections

Every section a sponsor needs to sign off

The skill drafts all 11 sections from what you give it. Anything unconfirmed gets [NEEDS GEMBA]. That's a to-do list, not a failure. Baseline metrics carry source tags so nobody argues with them on day 1.

§ 01
Business Case
A number tied to a real consequence: cost, safety, delivery, quality, capacity.
§ 02
Scope
Exact start and end points of the process. And what's explicitly out of scope.
§ 03
Boundaries & Constraints
Specific enough to test against a real proposal on day 3.
§ 04
Team (by role)
Sponsor, facilitator, team lead, members, SMEs. By function, not just names.
§ 05
Baseline Metrics
Every number has a source tag. No unsourced baseline.
§ 06
Targets
Specific, measurable targets tied to the business case metrics.
§ 07
Pre-Work Checklist
Complete enough that day 1 starts on the process, not on finding a stopwatch.
§ 08
Agenda (3-day or 5-day)
Chosen from your readiness answers, not defaulted without checking.
§ 09
Deliverables
What the team hands to the sponsor at the report-out.
§ 10
Report-Out Plan
Who presents, to whom, when, and what format.
§ 11
30/60/90-Day Follow-Up
Dates and an owner. A charter that ends at report-out has no way to know if the gain held.

Readiness First

The questions a good coach always asks before chartering

Two or more negative answers here means the event needs pre-work, or a root-cause pass before you charter at all. The skill names the risk explicitly rather than drafting around it.

👁️
Has anyone watched this process recently?
A week of redesign on a process nobody's observed is a week of guessing. Pre-work: schedule a gemba walk before chartering.
⚖️
Is the process stable enough to improve right now?
A kaizen event on an unstable process, such as one in crisis or mid-changeover to a new system, usually just standardizes the chaos. Fix the instability first.
🔓
Will the sponsor actually free the team for the full week?
No partial team, no pulling members back "just for an hour." Get a real commitment before drafting a charter that depends on it.
🎯
Is the problem understood well enough to scope a week around it?
If the root cause is still a guess, a kaizen event chases symptoms. A root-cause pass first often narrows scope and sharpens the target.

Why Kaizen Events Fail

The five failure modes this skill prevents

Business case that's really a complaint
"We've always wanted to fix this" isn't a business case. The skill asks for a number tied to a consequence: cost, safety, missed delivery, quality escapes.
Scope with no "out of scope" line
An invitation to scope creep on day 2. The charter names exactly what the team will not touch this week, and the skill flags any charter that's missing that line.
Team of "TBD"
A charter with unnamed roles can't verify freed-team commitments. The skill drafts roles even if names come later: sponsor, facilitator, team lead, members, SMEs by function.
Unsourced baseline gets argued on day 1
Every baseline metric in the charter carries a source tag. A number without one becomes a debate at the opening meeting instead of a starting point.
Follow-up that's "check back later"
A charter that ends at the report-out has no mechanism to know if the gain held at 30, 60, or 90 days. Dates and owners are required, not optional.

Agenda Formats

3-day or 5-day, chosen from your readiness answers

Default is 5-day when the process hasn't been mapped yet or when there's no baseline data in hand. 3-day when the current-state map exists and the week is mainly design-and-test.

5-Day Event
3-Day Event
Day 1
Observe & Map

Gemba walk, current-state VSM, time studies, waste identification. Team aligns on what they actually saw, not what they were told.

Day 2
Analyze & Ideate

Root cause on the biggest losses, future-state VSM, countermeasure brainstorm. Prioritize by impact vs. effort.

Day 3
Design & Pilot

Build the future-state: rearrange, remove, combine steps. First pilot run with the team watching.

Day 4
Test & Iterate

Multiple pilot cycles with real production. Document the new standard. Identify what needs sustained.

Day 5
Report-Out & Handoff

Before/after data, standard work drafted, 30/60/90-day action plan, sponsor sign-off. Celebrate.

Day 1
Confirm Current State & Design

Quick gemba walk to validate the existing current-state map, then move directly into future-state design and prioritized countermeasures.

Day 2
Pilot & Iterate

Build and test the future-state with real work. Iterate until the cycle runs smoothly. Document as you go.

Day 3
Stabilize & Report

Lock in the standard, capture before/after metrics, present to sponsor, assign 30/60/90-day owners.

What you get

A real run on the sample data

Unedited output. The sample message at the top went to Claude Sonnet 4.6 (an earlier-generation model, so a newer one may word things differently) with this skill pasted in, the same way the steps below show, on September 29, 2026. Scroll inside the frame to read the whole reply.

Kaizen Charter: the full reply from a real run on the sample data, ending with what the skill did, what still needs a human, and one next step.

Open the full image

Installation

Charter in one session

1
Copy the skill. Open kaizen-charter.md, select all the text and copy it.
2
Paste it into Claude. In claude.ai, start a new chat and paste it as your first message, or paste it into a Project's instructions so it's there every time. The free plan works, and free accounts can keep up to five Projects.
3
Say what's in front of you, in your own words, with any data you have (leave out names) such as "we're planning a kaizen event for [process]." Add "just draft it" to skip the readiness questions and get the whole charter now.
4
See a finished example first (optional): the worked examples are in the download, in the resources/examples folder.
5
Export to Word (optional): in Claude Code, the skill can also export a Word file with scripts/export_docx.py. In a plain chat, just copy the answer out.
Draft-first mode

Say "just draft it" at any point and the skill stops coaching and produces the full charter immediately, with readiness risks named in a dedicated section, not silently omitted.

Related Skills

Before and after a kaizen charter

🗺️ VSM Calculator: scope the event around bottleneck data 📝 TWI Job Breakdown Sheet: train the new standard after the event 🧭 Start Here: not sure if a kaizen event is the right move?