Intake
Intake is the system for collecting the right information, in the right order,
with the right level of trust, before real work begins.
Trigger Conditions
Use this skill when the user needs to:
- - onboard a new client, patient, customer, tenant, student, or employee
- design an intake form or discovery questionnaire
- structure an onboarding workflow
- gather requirements before providing a service
- identify what information should be collected at the start of a relationship
- improve an existing intake process that feels incomplete, bureaucratic, or low-conversion
Also trigger when the user says things like:
- - "I need an intake form"
- "Help me onboard a new client"
- "What should I ask before starting"
- "I need a discovery process"
- "How do I collect the right information upfront"
The Information You Wish You Had Asked
Every service relationship has a moment — usually six weeks in — where the provider thinks:
I wish I had known that at the start.
The therapist who discovers the client has tried this approach twice before.
The contractor who finds out the client had a bad experience with the last three contractors.
The accountant who learns about the offshore account in month four.
The coach who realizes the presenting goal is not the real goal.
Intake is not paperwork. It is the intelligence-gathering operation that determines whether
the work that follows will be effective or expensive.
This skill builds intake systems that surface the information that matters — before the
work begins.
Core Principle
People do not withhold important information because they are secretive. They withhold it
because nobody asked the right question, or because the form felt like bureaucracy rather
than care. The intake process that feels like a thoughtful conversation extracts more
useful information than the intake process that feels like a legal requirement.
Design for the conversation. The information follows.
Default Outputs
Depending on the request, produce one or more of the following:
- 1. Intake Form
A structured form with sections, field names, and question wording.
- 2. Intake Interview Guide
A conversation-based script for live intake calls or first sessions.
- 3. Intake Workflow
A step-by-step onboarding flow showing what is collected, when, and why.
- 4. Intake Summary Template
A reusable template for summarizing what was learned during intake.
- 5. Intake Audit
A diagnostic review of an existing intake process, including friction points,
missing questions, over-collection, and compliance considerations.
Response Rules
When responding:
- - first identify the service type
- separate administrative intake from discovery intake
- avoid collecting information that will not be used
- explain why sensitive questions are asked
- prefer clarity and usability over comprehensiveness
- flag privacy or regulatory issues without pretending to give legal, medical, or financial advice
Intake System Architecture
~~~python
INTAKE_FRAMEWORK = {
"information_layers": {
"layer
1administrative": {
"purpose": "Legal, billing, and contact requirements",
"examples": ["Full legal name", "Date of birth", "Contact information",
"Insurance or billing details", "Emergency contact",
"Consent and authorization signatures"],
"format": "Form — this is genuinely administrative, forms are appropriate here"
},
"layer
2presenting": {
"purpose": "Why they are here and what they want",
"examples": ["Primary reason for seeking service",
"What they hope to achieve",
"What has already been tried",
"Timeline and urgency"],
"format": "Questionnaire with open-ended fields — not multiple choice"
},
"layer
3context": {
"purpose": "The information that changes how you provide the service",
"examples": ["Relevant history", "Previous experiences with similar services",
"Factors that affect delivery", "Preferences and constraints",
"What success looks like to them specifically"],
"format": "Intake interview — conversation extracts what forms miss"
},
"layer
4relationship": {
"purpose": "How to work with this person effectively",
"examples": ["Communication preferences", "Decision-making style",
"What they need from a provider relationship",
"Previous experiences that shaped their expectations"],
"format": "Discovery conversation — built into first session or call"
}
}
}
~~~
Intake Form Design
~~~python
FORM_DESIGN = {
"principles": {
"minimum
viableintake": "Collect only what you will actually use — every field
not used creates friction without return",
"progressive_disclosure": "Administrative fields first, sensitive questions later —
trust is built before vulnerability is requested",
"open
fieldsover_checkboxes": "Checkboxes tell you what you gave them to say.
Open fields tell you what they actually think.",
"explain_why": "A sentence explaining why you need sensitive information
dramatically increases completion rate and accuracy"
},
"question_design": {
"weak": "Do you have any medical conditions? [ ] Yes [ ] No",
"strong": "Please describe any physical conditions, injuries, or health factors
that are relevant to our work together. This helps me tailor my approach
from our first session.",
"weak_2": "What are your goals?",
"strong_2": "If our work together goes exactly as you hope, what will be different
in your life six months from now? Be as specific as you can."
},
"completionratefactors": [
"Mobile-optimized — 60%+ of forms are completed on phone",
"Progress indicator for multi-page forms",
"Save and resume for longer forms",
"Plain language — no jargon or clinical terminology",
"Estimated completion time stated upfront",
"Confirmation message that feels human not automated"
]
}
~~~
The Intake Interview
~~~python
INTAKE_INTERVIEW = {
"structure": {
"opening": {
"purpose": "Establish safety and set expectations",
"script": "Before we dive in, I want to explain how I use what you share with me.
Everything you tell me today helps me [specific benefit]. There are no
wrong answers — the more honest you can be, the better I can help you."
},
"presenting": {
"questions": ["What brings you here now — why today rather than six months ago",
"What have you already tried, and what happened",
"What would have to change for you to feel this was worth it"],
"listen_for": "Gap between stated goal and real goal — they are often different"
},
"history": {
"questions": ["What is the relevant background I should know",
"Have you worked with someone on this before — what was that like",
"What has gotten in the way in the past"],
"purpose": "Previous failed attempts contain more useful information than successes"
},
"expectations": {
"questions": ["What do you expect this process to look like",
"What would make you want to stop working together",
"What do you need from me that you have not always gotten"],
"purpose": "Surface misaligned expectations before they become relationship problems"
},
"closing": {
"script": "Is there anything important that I have not asked about — something
you came in hoping I would understand that we have not covered yet",
"purpose": "This single question surfaces more useful information than
most of the rest of the interview combined"
}
},
"activelisteningprotocol": """
INTAKE_LISTENING = {
"notewhatis_avoided": "Topics skirted or answered too quickly often matter most",
"noteenergyshifts": "Where does engagement increase or decrease",
"note_contradictions": "Between stated priorities and described behavior",
"note_language": "Their exact words for their problem become your vocabulary",
"donotsolve_yet": "Intake is information gathering — resist the urge to help
before you understand"
}
"""
}
~~~
Intake by Service Type
~~~python
SERVICE_SPECIFIC = {
"healthcare_therapy": {
"required": ["Chief complaint", "Medical history", "Medications and allergies",
"Previous treatment", "Family history if relevant",
"Consent and privacy acknowledgment"],
"sensitive": "Trauma history, substance use, mental health history —
ask with care, explain purpose, never require before trust is established"
},
"legal": {
"required": ["Matter description", "Relevant dates and deadlines",
"Prior legal representation", "Documents in possession",
"Desired outcome", "Budget and timeline expectations"],
"conflict_check": "Run conflict check before intake interview — before sensitive
information is shared"
},
"financial_advisory": {
"required": ["Current financial situation", "Income and assets",
"Liabilities and obligations", "Risk tolerance",
"Goals and timeline", "Previous advisory relationships"],
"sensitive": "Financial shame is real — create psychological safety before
asking for numbers"
},
"coaching": {
"required": ["Presenting goal", "Current situation", "Previous attempts",
"Resources available", "Accountability preferences",
"Definition of success"],
"distinguish": "Presenting goal vs real goal vs root cause —
coaching intake should surface all three"
},
"creative_services": {
"required": ["Project scope and deliverables", "Timeline and milestones",
"Budget", "Decision-making process", "Brand guidelines",
"Examples of work they love and hate — both equally important"],
"critical": "Examples of work they hate prevent the most expensive revisions"
}
}
~~~
Information Management
~~~python
INTAKE
DATAMANAGEMENT = {
"storage_principles": [
"Information collected for one purpose should not be used for another",
"Sensitive information needs appropriate security — not in email threads",
"Retention policy defined before collection begins",
"Destruction process for data when relationship ends"
],
"usability": {
"principle": "Intake information only has value if it is used",
"system": "Review intake notes before every session — not just the first",
"update": "Intake is not a one-time event — update as circumstances change",
"transfer": "If client transfers to another provider, intake summary
is the most valuable thing you can provide"
},
"compliance_flags": {
"healthcare": "HIPAA (US) or equivalent — specific requirements for PHI",
"financial": "KYC and AML obligations for financial services",
"general": "GDPR, CCPA, or applicable privacy regulation — get legal advice
for your jurisdiction before collecting sensitive data"
}
}
~~~
Boundaries
This skill helps design intake systems, forms, interviews, and information workflows.
It does not replace legal, medical, financial, privacy, or regulatory advice.
For regulated contexts, adapt outputs to the user's jurisdiction and consult a qualified
professional before using the workflow in production.
Quality Check Before Delivering
- - [ ] Information layers separated — administrative, presenting, context, relationship
- [ ] Form questions are open-ended not checkbox-dominated
- [ ] Sensitive questions explained with purpose
- [ ] Intake interview includes closing question — what have I not asked
- [ ] Service type matched to specific intake requirements
- [ ] Data storage and compliance flagged for sensitive service types
- [ ] Review protocol built in — intake used throughout relationship not just at start
信息采集
信息采集是在实际工作开始前,以正确的顺序和适当的信任级别收集正确信息的系统。
触发条件
当用户需要以下内容时使用此技能:
- - 为新客户、患者、顾客、租户、学生或员工办理入职
- 设计信息采集表或发现问卷
- 构建入职工作流程
- 在提供服务前收集需求
- 确定在关系建立之初应收集哪些信息
- 改进现有感觉不完整、官僚化或转化率低的信息采集流程
当用户说出类似以下内容时也触发:
- - 我需要一份信息采集表
- 帮我为新客户办理入职
- 开始前我应该问什么
- 我需要一个发现流程
- 如何提前收集正确的信息
你希望当初问过的问题
每个服务关系都有一个时刻——通常是六周左右——服务提供者会想:
我当初要是知道这个就好了。
治疗师发现客户以前尝试过这种方法两次。
承包商发现客户对前三家承包商有过糟糕经历。
会计师在第四个月才知道海外账户的事。
教练意识到表面目标并非真正目标。
信息采集不是文书工作。它是情报收集行动,决定了后续工作是否有效还是代价高昂。
此技能构建的信息采集系统能在工作开始前揭示重要信息。
核心原则
人们隐瞒重要信息并非因为守口如瓶。他们隐瞒是因为没人问对问题,或者因为表格感觉像是官僚程序而非关怀。感觉像是一次深思熟虑对话的信息采集过程,比感觉像是法律要求的信息采集过程能提取更多有用信息。
为对话而设计。信息自然随之而来。
默认输出
根据请求,生成以下一项或多项:
- 1. 信息采集表
包含章节、字段名称和问题措辞的结构化表格。
- 2. 信息采集访谈指南
用于实时信息采集通话或首次会面的基于对话的脚本。
- 3. 信息采集工作流程
分步入职流程,显示收集什么、何时收集以及为何收集。
- 4. 信息采集摘要模板
用于总结信息采集过程中所学内容的可复用模板。
- 5. 信息采集审计
对现有信息采集流程的诊断审查,包括摩擦点、缺失问题、过度收集和合规考量。
响应规则
回应时:
- - 首先确定服务类型
- 将行政信息采集与发现信息采集分开
- 避免收集不会被使用的信息
- 解释为何询问敏感问题
- 优先考虑清晰度和可用性而非全面性
- 标记隐私或监管问题,但不假装提供法律、医疗或财务建议
信息采集系统架构
~~~python
INTAKE_FRAMEWORK = {
information_layers: {
layer
1administrative: {
purpose: 法律、计费和联系要求,
examples: [法定全名, 出生日期, 联系方式,
保险或计费详情, 紧急联系人,
同意和授权签名],
format: 表格——这确实是行政性的,此处表格适用
},
layer
2presenting: {
purpose: 他们为何而来以及他们想要什么,
examples: [寻求服务的主要原因,
他们希望实现什么,
已经尝试过什么,
时间线和紧迫性],
format: 带有开放式字段的问卷——而非选择题
},
layer
3context: {
purpose: 改变你提供服务方式的信息,
examples: [相关历史, 类似服务的先前经历,
影响交付的因素, 偏好和限制,
对他们而言成功是什么样子],
format: 信息采集访谈——对话能提取表格遗漏的内容
},
layer
4relationship: {
purpose: 如何有效与这个人合作,
examples: [沟通偏好, 决策风格,
他们需要从服务提供者关系中获得什么,
塑造他们期望的先前经历],
format: 发现对话——融入首次会面或通话中
}
}
}
~~~
信息采集表设计
~~~python
FORM_DESIGN = {
principles: {
minimum
viableintake: 只收集你实际会使用的信息——每个未使用的字段都会产生摩擦而无回报,
progressive_disclosure: 先行政字段,后敏感问题——信任建立后才要求脆弱性,
open
fieldsover_checkboxes: 复选框告诉你他们说了你给他们的选项。开放式字段告诉你他们真实的想法。,
explain_why: 解释为何需要敏感信息的一句话能显著提高完成率和准确性
},
question_design: {
weak: 您有任何医疗状况吗?[ ] 是 [ ] 否,
strong: 请描述任何与我们合作相关的身体状况、损伤或健康因素。这有助于我从首次会面就调整方法。,
weak_2: 您的目标是什么?,
strong_2: 如果我们的合作完全如您所愿,六个月后您的生活会有什么不同?请尽可能具体。
},
completionratefactors: [
移动端优化——60%以上的表格在手机上完成,
多页表格的进度指示器,
较长表格的保存和继续功能,
通俗语言——无行话或临床术语,
提前说明预计完成时间,
感觉人性化而非自动化的确认信息
]
}
~~~
信息采集访谈
~~~python
INTAKE_INTERVIEW = {
structure: {
opening: {
purpose: 建立安全感并设定期望,
script: 在深入之前,我想解释我如何使用您分享的信息。您今天告诉我的每件事都有助于我[具体好处]。没有错误答案——您越诚实,我越能帮助您。
},
presenting: {
questions: [您现在为什么来——为什么是今天而不是六个月前,
您已经尝试过什么,结果如何,
需要改变什么您才会觉得这值得],
listen_for: 陈述目标与真实目标之间的差距——它们往往不同
},
history: {
questions: [我应该了解哪些相关背景,
您以前和谁合作过这个问题——感觉如何,
过去是什么阻碍了您],
purpose: 过去的失败尝试比成功包含更多有用信息
},
expectations: {
questions: [您期望这个过程是什么样子,
什么会让您想停止合作,
您需要从我这里得到什么而您并不总能得到],
purpose: 在期望不一致变成关系问题之前将其揭示
},
closing: {
script: 有没有什么重要的事情我没有问到——您来的时候希望我理解但我们还没有涵盖的内容,
purpose: 这一个问题揭示的有用信息比访谈其余部分加起来还多
}
},
activelisteningprotocol:
INTAKE_LISTENING = {
notewhatis_avoided: 被回避或回答过快的话题往往最重要,
noteenergyshifts: 参与度在哪里增加或减少,
note_contradictions: 在陈述的优先事项和描述的行为之间,
note_language: 他们描述问题的原话成为你的词汇,
donotsolve_yet: 信息采集是信息收集——在理解之前克制帮助的冲动
}
}
~~~
按服务类型的信息采集
~~~python
SERVICE_SPECIFIC = {
healthcare_therapy: {
required: [主诉, 病史, 药物和过敏史,
既往治疗, 相关家族史,
同意和隐私确认],
sensitive: 创伤史、物质使用、心理健康史——谨慎询问,解释目的,在信任建立前绝不要求
},
legal: {
required: [案件描述, 相关日期和截止日期,
先前法律代理, 持有的文件,
期望结果, 预算和时间期望],
conflict_check: 在信息采集访谈前进行利益冲突检查——在分享敏感信息之前
},
financial_advisory: {
required: [当前财务状况, 收入和资产,
负债和义务, 风险承受能力,
目标和时间线, 先前咨询关系],
sensitive: 财务羞耻感真实存在——在询问数字前创造心理安全感
},
coaching: {
required: [表面目标, 当前状况, 先前尝试,
可用资源, 问责偏好,
成功的定义],
distinguish: 表面目标 vs 真实目标 vs 根本原因——教练信息采集应揭示三者
},
creative_services: {
required: [项目范围和交付物, 时间线和里程碑,
预算, 决策流程, 品牌指南,
他们喜欢和不喜欢的作品示例——两者同等重要],
critical: 他们不喜欢的作品示例能防止最昂贵的修改
}
}
~~~
信息管理
~~~python
INTAKE
DATAMANAGEMENT = {
storage_principles: