🤝 The project kickoff ritual
Every consulting project starts the same way: a kickoff meeting, a bunch of questions, a list of access requests, and a schedule to agree on. It’s work repetitive, easy to standardize, and still mostly done by hand. The onboarding skill exists to compress this ritual into minutes: describe the client, and it builds the entire package.
🎯 What onboarding needs to deliver
A professional onboarding package covers five things—the five pillars of any kickoff:
- •Scope — what is inside and outside the project.
- •Stakeholders — who decides, who approves, who grants access.
- •Access — credentials, accounts, and permissions to collect.
- •Expectations — what success looks like, and how long it takes.
- •Timeline — the first week, milestones, and deliverables.
💡 Why automating this is so worthwhile
Onboarding is the consultant’s most predictable deliverable—and that’s why it’s the most wasteful when done manually. Every new project spends the same hours rewriting the same documents. A skill that produces the package standardizes your delivery, cuts the time from “sold” to “started working,” and gives the client a first impression of organization that no standalone email can.
📋 Anatomy of the onboarding package
The skill doesn’t return generic text—it returns named artifacts, each with a role in the kickoff. Knowing each one lets you adapt the structure to your niche without reinventing anything.
Welcome doc
First formal contact
Welcome, summary of the agreed scope, how we'll work together, and what to expect over the coming weeks. Sets the tone for the relationship.
Discovery questionnaire
Context gathering
The questions that draw out what you need to know: current processes, pain points, available data, and goals. This is the input that feeds the audit report later.
Access checklist
What to ask for and when
List of credentials and permissions — CRM, analytics, repositories, cloud accounts — with an owner and deadline. Prevents the classic bottleneck of “we’re waiting for access.”
Week 1 plan
Initial timeline
First-days milestones: meetings, access collection, start of the diagnosis. Turn kickoff enthusiasm into concrete action.
🧩 Practical tip
Each artifact is a template with placeholders ({{cliente}}, {{escopo}}, {{stack}}). When you write your version of the skill, just replace the text with your own — the engineering is in the structure, not the wording.
🔎 From Diagnosis to AI Readiness Report
The second skill is where the consulting delivers value. You leave the meeting with raw notes — loose observations, client phrases, problems spotted. The skill turns this into a AI Readiness Report: a report with a readiness score, structured findings, and a prioritized roadmap.
✓ A good Readiness Report
- ✓Gives an objective score for each dimension, not an opinion.
- ✓Connects each finding to impact and effort.
- ✓Ends with a roadmap the client can approve.
- ✓Highlights quick wins with immediate impact.
✗ A weak report
- ✗Lists problems without prioritizing them.
- ✗Recommends “implement AI” without saying where or how.
- ✗Ignores the effort and promises the impossible.
- ✗Doesn’t connect the finding to a concrete action.
📐 The dimensions of AI readiness
What makes a report a reusable framework is evaluating the company on fixed axes, and not by intuition. Each axis gets a score and produces findings. That makes the same report work for a store, a clinic, or a SaaS.
| Dimension | What is evaluated | Low-readiness signal |
|---|---|---|
| Data | Where they are, format, quality, access. | Scattered spreadsheets, no single source of truth. |
| Processes | What is manual, repetitive, and documented. | Everything in one person's head. |
| Team | AI familiarity, appetite for change. | Generalized fear or skepticism. |
| Infrastructure | Tools, integrations, automations. | Isolated systems, no API. |
| Use cases | Where AI would deliver clear returns. | No obvious candidates mapped. |
📊 How the score guides the conversation
- High data maturity + low process maturity → start by automating workflows; the data is already ready.
- Curious team + weak infrastructure → invest in integration before any model.
- Clear use cases + bad data → the first project is to clean up the data, and that needs to be stated.
⚙️ How the two skills fit together
Separately, they’re two utilities. Chained together, they become a consulting product. The onboarding opens the project and gathers context; the Audit Report closes the diagnosis using exactly that context. One’s output is the other’s input.
🔗 The consulting pipeline
welcome · questionnaire · access · week 1 plan
↓ questionnaire responses
[ Audit Report skill ]
notes + answers → score · findings · roadmap
↓
[ Project proposal ]
🧩 Practical tip: shared context
Save the questionnaire answers in a file that both skills read (e.g., cliente.md). That way, the Audit Report doesn't ask again for what onboarding already collected — the handoff stays clean, and the client doesn't answer the same question twice.
🛠️ Building your version
The generic skill is the starting point. What sets it apart is its specialization: your fields, your tone, your brand. Below is the skeleton of a SKILL.md that generates the onboarding package and report — ready for you to adapt.
--- name: ai-consulting-onboarding description: > Gera um pacote completo de onboarding de cliente e um AI Readiness Report para projetos de consultoria em IA. Dispara quando o usuario diz "novo cliente", "kickoff", "pacote de onboarding", "relatorio de prontidao para IA", "audit report" ou descreve uma empresa a avaliar. --- # AI Consulting Onboarding + Audit Report Voce e um consultor de IA. Produza entregaveis reais, nao conselhos. ## Fase 1 — Onboarding Pergunte (ou leia de cliente.md): nome, setor, escopo, stakeholders, stack atual, objetivo de 90 dias. Gere os arquivos: - welcome.md (boas-vindas + escopo) - questionario.md (perguntas de descoberta) - acessos.md (checklist de credenciais) - plano-semana-1.md (cronograma inicial) ## Fase 2 — AI Readiness Report A partir das notas + respostas, avalie 5 dimensoes (dados, processos, equipe, infraestrutura, casos de uso). Para cada uma: nota 1-5 + achado. Gere readiness-report.md com: - Sumario executivo (3 bullets) - Tabela de notas por dimensao - Achados (impacto x esforco) - Roadmap priorizado (quick wins / medio / longo prazo) ## Referencias - references/dimensoes.md (rubrica de cada eixo) - references/templates.md (modelos dos 4 documentos)
⚡ Prompt to trigger the skill
📝 Module Summary
🎯 Hands-on exercises
- Write the 4 onboarding templates (welcome, questionnaire, access, plan) for the your niche, with placeholders.
- Define the 5-dimension rubric: what earns a score of 1 and what earns a score of 5 on each axis.
- Create a runnable SKILL.md called
ai-consulting-onboardingthat generates the package and report from a company description. Test it with a fictional client and check that the files are named and filled in. - Add a
references/dimensoes.mdwith the rubric and have the SKILL.md read it on demand (progressive disclosure).
Next Module:
5.2 — SEO + AI Search Auditor