You have worked the five modules — the foundation, the hire, onboard & develop, reviews & retention, and org change. Each one ended with a graded artifact: the competency framework, the job description, and the sensitive communications template; the structured scorecard, the panel synthesis, and the offer and rejection letters; the onboarding plan and the policy rewrite; the performance review, the engagement action plan, and the offboarding checklist; the org change narrative in three versions, the rollout sequence, and the feedback loop. The capstone is where those five stop being five.
This is the integrative project that earns the credential. One company’s people function, proven by running it through one full growth quarter end to end — a complete HR operating system where every decision traces back to the same competency framework, in English and Arabic, from the job description to the all-hands change announcement. That is the thing “Certified People & HR with Claude” attests, and it is a higher bar than five modules completed: a certified people operator is not someone who watched the track, it is someone who can run a whole quarter’s people function on a system that coheres and survives scrutiny from the team, a lawyer, and the board.
The capstone is not five artifacts stapled together. It is one test: does the whole people function hold as a single system — where the competency framework you built in Module 1 is still load-bearing in the performance review in Module 4 and the org change narrative in Module 5, and every decision traces all the way back to the same observable behaviors? A reviewer should not find a hiring decision that contradicts the framework it claims to be anchored to. Integration is the grade — coherence across the whole system, not five separate deliverables that each look fine alone.
The brief
Build a complete People & HR operating system for one company, and prove it by running its people function through one full growth quarter end to end. We recommend your own org — the capstone then doubles as a real quarter’s HR work and infrastructure you keep (see Bring your own org) — or the alternative, Mizan’s Q2 2026, the GCC bookkeeping SaaS whose people team is threaded through the whole track, if you would rather learn on neutral ground.
You already know the Desktop motion: open the company’s folder in Claude Desktop, approve each read in the “Ask permissions” prompt, and assemble the system in the chat and the file pane. No terminal on this path. Claude drafts the frameworks and the narratives fast; you check every judgment call, name the decisions a person had to make, and ensure sensitive data — compensation, medical, personal — stays inside the workspace.
Capstone deliverable — People-in-a-Box (one company, one quarter, end to end)
1. The foundation (Foundation / M1)
- competency-framework.md — one role family, observable behaviors per level
- inclusive-jd.md — genuine must-haves only, bilingual, bias-checked
- sensitive-comms.md — one hard message, decision first, no burying the lede
2. The hiring kit (Hire / M2)
- structured scorecard anchored to the M1 framework
- panel synthesis — the data/style distinction named, dissent surfaced
- offer letter + two specific rejection letters (not "went another direction")
3. Onboard & develop (Onboard / M3)
- onboarding-plan.md — week-by-week, milestone on Day 30, manager checkpoints
- policy-rewrite.md — one opaque policy rewritten into one page of rules
4. Reviews & retention (Review / M4)
- performance-review.md — competency-anchored, bias-checked, growth signal named
- engagement-action-plan.md — two commitments with dates, one honest acknowledgment
- offboarding-checklist.md — knowledge captured before last day, access in order
5. Org change (Change / M5)
- org-change-narrative.md — three versions (1:1 first-told, manager pack, all-hands)
- rollout-sequence.md — who hears what, in which order, and why
- feedback-loop.md — one question per manager, results shared back to the team
6. A short reflection
- where Claude genuinely helped, and where your judgment was needed
- the ONE decision you made differently because a person had to
Every piece must trace to #1. That traceability is the whole assignment.The reflection matters: naming where you overrode the draft — the performance review feedback you rewrote because the first version was too soft, the change narrative you re-sequenced because the draft missed the right first audience, the policy rule you flagged for legal before shipping — is part of the proof that you ran Claude well rather than letting it run you.
How it’s assessed — the master rubric
Five criteria, each scored meets / nearly / not yet. One “nearly” anywhere is a revise, not a pass — because a people function that coheres in four places and drifts in the fifth still has one decision nobody can defend, and an undefendable decision is exactly what fails later: in a grievance, in a legal review, in the conversation with the person it affected. The bar throughout is a single question: would a chief people officer be able to defend every decision in this system to the team, to a lawyer, and to the board?
Master rubric — People-in-a-Box
Each criterion: meets / nearly / not yet. The bar: would a chief people officer
be able to defend every decision in this system to the team, to a lawyer,
and to the board?
1. Competency is the through-line
meets The M1 framework is still load-bearing in the M2 scorecard,
the M4 review, and the M5 change narrative. A stranger can
see the same observable behaviors being measured, reviewed,
and referenced across the system.
nearly The framework exists but loosens downstream — the scorecard
adds criteria not in the framework, or the review uses
language the framework doesn't.
not yet Each piece uses its own implicit criteria. The framework
was built once and then ignored.
2. Decisions are human-owned
meets No hiring decision, review rating, engagement action, or
change announcement was shipped as raw model output. Every
one has a named human who decided and signed; Claude
drafted, a person led.
nearly Most decisions are owned, but one slipped through unsigned
or unchecked, or the reflection asserts judgment without
showing it.
not yet Reads as model output — confident, unreviewed, no named
owner on the decisions that affect people's careers.
3. The fairness gates held
meets Bias was explicitly checked in at least the performance
review and the org change narrative; anonymity was protected
in the engagement action plan; sensitive data (compensation,
medical, personal) stayed in the workspace.
nearly Most gates held, but one sensitive item escaped the
workspace, or a bias check was skipped on a high-stakes
artifact.
not yet No explicit bias check visible; sensitive data handled
without controls; the fairness layer is absent.
4. Arabic is first-class
meets The JD, the all-hands change announcement, and the
performance review exist in Arabic, authored (not
translated), at Gulf-natural register.
nearly Competent but translation-flavored, or the register slips,
or one of the three required artifacts is missing.
not yet English run through translation; reads foreign; or the
Arabic pieces were skipped.
5. The system coheres
meets All five parts are connected. A reviewer can trace from the
competency framework to the hiring decision to the
onboarding milestone to the first review signal to the
change narrative. No piece is a standalone artifact that
contradicts or ignores the others.
nearly A part is missing, or present but disconnected — a review
that doesn't reference the framework, a change narrative
that can't explain why this person was chosen.
not yet A collection of deliverables, not a people function.Notice that three of the five criteria — competency as through-line, decisions are human-owned, and the system coheres — grade the seams between the pieces, not the quality of any one of them alone. That is intentional. The modules already graded each artifact on its own; the capstone grades whether the framework, the scorecard, the onboarding plan, the review, and the change narrative hold together as one people function across a quarter.
Bring your own org
Use your own org. We recommend it for one blunt reason: the capstone stops being coursework and becomes two things at once — a quarter’s people work you actually ran, and infrastructure you keep. The system you assemble — the competency framework, the hiring kit, the onboarding plan, the performance review and engagement habits, the change management templates, in both languages — is real infrastructure your HR function keeps and runs, and the quarter you run it on is work you actually ship. That is day-one ROI: the assessment, the portfolio piece, and the actual work are the same artifact.
The sample — Mizan’s Q2 2026 — is here for when neutral ground is the point: an HR generalist who wants to practice the integration once before doing it on live headcount, or a team that wants a worked reference before rolling this across the org. Mizan’s Q2 is worked through every module — the CS competency framework built in M1, the Senior CSM hire in M2, Rania’s onboarding in M3, Youssef’s review and Khalid’s offboarding in M4, and the CX restructuring in M5 — so you can study the moves with no career consequences attached, then turn around and do it for your own org. Either path clears the same rubric.
What you earn
Clear the rubric and you earn “Certified People & HR with Claude” — a verifiable page at /cert/[id] you can share and add to LinkedIn. It is the credential of the team track: proof that you can run a whole quarter’s people function on a coherent system, not just describe the moves.
Here is the honest state of it today, because the credential is only worth what it honestly claims:
- A person grades it, now. In a cohort or concierge run, a reviewer scores your HR system against the master rubric above and tells you what is weak. The rubric and the chief-people-officer bar are the IP, and they are hand-gradeable today — that is the deliberate first step, not a placeholder.
- AI-assisted grading on the same rubric is next, not a claim we make yet. When it ships, it grades against this exact rubric — the bar does not move, the reviewer does.
- It is not an accreditation-body badge. Its weight is the work behind it: a complete, coherent people function a reviewer judged against a chief-people-officer bar. That is a real thing to point at — and a more useful one than a quiz score.
No automated verification or paywall is wired behind this today that the line above does not describe. The certificate means what the work means.
What good looks like
The detailed examples live in the five modules. What the capstone adds is the through-line — the same competency framework visibly threading every decision across one quarter. Here is Mizan’s Q2 People-in-a-Box at a glance, one or two lines per stage, so you can see the inheritance the rubric is really grading:
Mizan — People-in-a-Box at a glance (Q2 2026)
Through-line The CS competency framework (M1) is the document every
downstream decision traces to. The Senior CSM level's
"observable behavior" for client management is what the
interview scorecard graded, what the 30-day milestone
tested, and what the org change cited when asked why
Nadia was the right Head of CX.
Foundation competency-framework.md (CS function, 3 levels),
inclusive-jd.md (Senior CSM — genuine must-haves only,
bilingual required, GCC experience preferred not required),
sensitive-comms.md (CEO → ops team on structural change:
decision first, uncertainty named, no burying the lede).
The hire Rania Al-Khaldi: scorecard anchored to the CS framework,
4-person panel, synthesis surfaced the data/metrics gap
and named the style-vs-signal distinction. Offer traced
to the scorecard. Two runners-up received specific
rejections — not "went another direction."
Onboard Rania's 30-day plan: Week 1 no solo client calls;
Day 30 milestone: lead one onboarding call, present
30-day reflection to Omar. Policy rewrite: Mizan client
data policy from 4 pages of jargon to one page of rules
("client data stays in Mizan — full stop").
Reviews Youssef review: Technical execution Meets, client
communication Developing. Growth signal: one specific
behavior by Q3. Bias check: style vs competency gap
distinguished explicitly. Engagement plan: career path
framework published by Q3-end (commitment), month-end
crunch acknowledged with earlier client data runway.
Khalid offboarding: integration knowledge captured in
one page before last day; access revoked in the right
order.
Org change CX restructuring: Nadia told first (1:1), then the two
generalists who applied (also 1:1), then managers, then
all-hands. The all-hands narrative led with the client
confusion problem — not the org chart. Manager pack
answered the hard question ("why Nadia?") with a
specific, honest answer. Two-week feedback loop: one
question per manager, results shared back to the team.Read it top to bottom and the test is obvious: the competency framework from Day 1 is still the answer to “why Nadia?” on Day 90. That is what “meets” on coherence looks like — not five strong deliverables, but one people system where the same observable behaviors that made the hiring decision defensible are what the performance review scored and what the org change cited. Nothing contradicts. That is the integration the capstone grades.
Now ship it
You have the brief, the rubric, and the bar. The capstone is not harder than the modules — it is the modules, run together for one quarter, with the seams made to hold. Pick the company, open the folder, and build the people system Claude drafted and you reviewed and owned.
One thing to keep beside you as you build: the competency framework from Module 1 is the document everything else loads from. If it is precise — observable behaviors, not trait words — every downstream piece has something real to anchor to. If it is vague, the whole system is built on air, and the capstone will show it. Start there, make it real, and the rest follows.
Build the system, clear the rubric, and the certificate is yours — because the work is.