fluency-assessment

AI Fluency Assessment

Purpose

Quick diagnostic that scores an organization's AI fluency across three pillars: psychological barriers, integration failures, and ownership gaps. Produces a one-page scorecard the leader can act on. This skill MUST run before any other interactive or workflow skill. It's the diagnostic before the prescription.

Core principle: Diagnose before you prescribe. This skill assesses — it does NOT build plans, pick tools, or write board narratives. Those are separate skills.

Time: Under 5 minutes for the quick quiz. Even faster if the leader has existing survey or assessment data. Optional deep-dive available if the leader wants more detail.

Flow

digraph fluency {
    "Context (2 Qs)" [shape=box];
    "Have existing data?" [shape=diamond];
    "Import & map data" [shape=box];
    "Gaps in data?" [shape=diamond];
    "Fill gaps (targeted Qs)" [shape=box];
    "Quick Quiz (9 Qs)" [shape=box];
    "Score & produce scorecard" [shape=box];
    "Offer team survey" [shape=box];
    "Want deep-dive?" [shape=diamond];
    "Deep-dive on weakest pillar" [shape=box];
    "Route to next skill" [shape=doublecircle];

    "Context (2 Qs)" -> "Have existing data?";
    "Have existing data?" -> "Import & map data" [label="yes"];
    "Have existing data?" -> "Quick Quiz (9 Qs)" [label="no"];
    "Import & map data" -> "Gaps in data?";
    "Gaps in data?" -> "Fill gaps (targeted Qs)" [label="yes"];
    "Gaps in data?" -> "Score & produce scorecard" [label="no"];
    "Fill gaps (targeted Qs)" -> "Score & produce scorecard";
    "Quick Quiz (9 Qs)" -> "Score & produce scorecard";
    "Score & produce scorecard" -> "Offer team survey";
    "Offer team survey" -> "Want deep-dive?";
    "Want deep-dive?" -> "Deep-dive on weakest pillar" [label="yes, have time"];
    "Want deep-dive?" -> "Route to next skill" [label="no, move on"];
    "Deep-dive on weakest pillar" -> "Route to next skill";
}

Process

1. Ask ONE question at a time. Never batch questions. 2. Wait for the leader's actual answer before proceeding. Never simulate or assume answers. 3. Present quiz questions with their multiple-choice options. The leader picks a letter. 4. Do NOT give advice, solutions, or recommendations during the assessment. Diagnose only. 5. Do NOT suggest tools, processes, or plans. That comes from other skills after this one.

Step 1: Context (2 questions)

Get just enough context to interpret the quiz answers. Ask one at a time:

1. "Tell me about your team — how big, what they do, and what's your role?"
2. "What AI tools does your team have access to today, if any?"

That's it. Don't ask about funding stage, runway, or history. Get to the assessment fast.

Step 2: Check for Existing Data

After context questions, ask:

"Have you already run a survey or assessment on AI adoption — even an informal one? If so, share the results and I'll map them to our framework instead of starting from scratch."

After the existing data question, also ask:

"Has your organization restructured, or is it planning to restructure, in connection with AI adoption — role changes, headcount adjustments, new positions, or team reorganization?"

If yes, ask one follow-up: "What changed — which roles, how many people affected?" A bare "yes" isn't enough — this follows the same principle as rejecting vague self-assessment.

Record the answer as one of three states for the scorecard:

• Stable — no restructuring
• Restructuring in progress — changes are happening now
• Restructuring completed [timeframe] — changes happened, team is now stable

If they have data: Go to Step 2a (Import Existing Data). If they don't: Go to Step 3 (Quick Quiz).

Step 2a: Import Existing Data

Accept whatever format they share — survey results, spreadsheet exports, summary notes, screenshots. Map their data to the three pillars using this approach:

1. Extract signals. Read through the data and identify anything that maps to psychological barriers, integration failures, or ownership gaps. Use the survey-to-pillar mapping table (see Team Survey Template section) as a guide, but be flexible — their questions won't match ours exactly.

2. Score what you can. For each pillar where you have enough signal, assign a 1-4 raw score based on what the data tells you. Use the same scoring logic as the quiz: 1 = no engagement, 4 = fluent.

3. Identify gaps. Check which pillars have weak or missing signal. Common gaps:

   • External surveys rarely cover psychological barriers well (people don't self-report fear or identity threat accurately)
   • Ownership gaps often aren't covered unless the survey explicitly asked about process and accountability
   • Integration data is usually the strongest since most surveys ask about tool usage

4. Fill gaps with targeted questions. For each pillar with insufficient data, ask 1-2 questions from the Quick Quiz (Step 3) that cover the gap. Tell the leader why:

   "Your survey gives me a good picture of [covered pillars]. I'm missing signal on [gap pillar] — let me ask [1-2] quick questions to fill that in."

   Pick the most diagnostic question per pillar:

   • Psychological Barriers gap → ask Q2 (reasons for non-use) and Q3 (leadership modeling)
   • Integration gap → ask Q4 (team relationship with tools) and Q6 (breadth of use)
   • Ownership gap → ask Q7 (who owns it) and Q8 (how you measure)

5. Proceed to scoring (Step 4) once all three pillars have signal.

Do NOT accept vague summaries as data. "Team seems positive" is not data. Push for specifics: "What percentage responded? What were the actual response distributions?" If the shared data is too vague to score, say so honestly and offer the quick quiz as a faster path than going back and forth on interpretation.

Step 3: Quick Quiz (9 questions)

Ask ONE question at a time. Present the options exactly as written. The leader picks A, B, C, or D. Each answer maps to a score (shown in brackets — do NOT show the scores to the leader).

Psychological Barriers (3 questions)

Q1. How does your team talk about AI tools?

• A) They don't — it doesn't come up [1]
• B) Mixed — some people are interested, others push back or avoid the topic [2]
• C) Generally positive, but a few vocal skeptics or quiet resisters [3]
• D) Most people are open to it — the conversation is about how, not whether [4]

Q2. When someone on your team says they don't use AI tools, what's the usual reason?

• A) "I don't need it" or "it's not good enough" — without having tried it seriously [1]
• B) "I tried it and it didn't help" — but only briefly or for the wrong task [2]
• C) "It doesn't fit how I work" — a specific workflow complaint [3]
• D) Most people use them — the holdouts are the exception, not the norm [4]

Q3. Do your leaders and managers use AI tools themselves?

• A) No — they talk about AI but don't use it personally [1]
• B) A couple do, but it's not visible to the team [2]
• C) Some do and share what they learn, but it's inconsistent [3]
• D) Yes — leadership models AI use and it's visible to the team [4]

Integration (3 questions)

Q4. How would you describe your team's current relationship with AI tools?

• A) We have licenses but most people haven't really started [1]
• B) Some individuals use them from time to time [2]
• C) A group uses them regularly for specific tasks [3]
• D) AI is part of how the team works — some tasks always start with AI now [4]

Q5. When people use AI, what does a typical interaction look like?

• A) Simple questions, take the answer as-is [1]
• B) They give some context to shape the output [2]
• C) They go back and forth, refining until the output is useful [3]
• D) They use templates, custom prompts, or built-in workflows for recurring tasks [4]

Q6. How many different tasks or workflow steps does your team use AI for?

• A) None, or just one thing (like autocomplete) [1]
• B) 2-3 things, but it's individual — everyone uses it differently [2]
• C) A few shared use cases that multiple people use [3]
• D) AI is used across several workflow stages — writing, reviewing, testing, docs, etc. [4]

Ownership (3 questions)

Q7. Who is responsible for making AI adoption work in your organization?

• A) Nobody specifically — or "me, alongside everything else" [1]
• B) It's loosely assigned to someone, but it's not their main focus [2]
• C) Someone has it as a real responsibility with some dedicated time [3]
• D) There's a named owner with a clear mandate, time, and authority [4]

Q8. How do you know if AI tools are actually helping?

• A) I don't — we have no way to measure it [1]
• B) We track license count or login activity [2]
• C) We have some usage data and anecdotal feedback [3]
• D) We measure specific outcomes — time saved, quality changes, or workflow metrics [4]

Q9. When someone has a problem with an AI tool, what happens?

• A) Nothing — they stop using it and nobody notices [1]
• B) They might mention it in passing, but there's no formal channel [2]
• C) There's someone they can ask, but no systematic feedback loop [3]
• D) There's a clear process — issues get reported, tracked, and addressed [4]

Step 4: Score

Scoring method:

• Each answer maps to a score of 1-4 (shown in brackets above)
• Per pillar: average the 3 question scores, then map to 1-5 scale:
  • Average 1.0-1.5 → Score 1/5
  • Average 1.6-2.2 → Score 2/5
  • Average 2.3-2.9 → Score 3/5
  • Average 3.0-3.6 → Score 4/5
  • Average 3.7-4.0 → Score 5/5
• Overall: average the three pillar scores

Scoring guidance:

• Score reflects the PROBLEM severity, not the team's quality. A score of 2/5 on psychological barriers means "significant unaddressed psychological resistance."
• Be honest. Leaders need accurate diagnosis, not encouragement.
• Higher score = better fluency (fewer barriers, better integration, stronger ownership).
• Score behavior, not access. Having licenses is not adoption. People logging in is not adoption. Adoption means the work has actually changed.

Produce the scorecard using the Output format below.

Step 5: Offer Team Survey + Optional Deep-Dive

After presenting the scorecard:

"This scorecard is based on your perspective — under 5 minutes, so it's a quick read. Two things I'd recommend:

1. Send a 5-minute survey to your team to get their side of the picture. I can draft it for you.
2. If you have 10 more minutes, I can do a deeper dive on your weakest pillar — [name it] — to understand the specific blockers.

Which would you like? Both? One? Or just move to the next step?"

If they want the deep-dive: Use the Deep-Dive Probes section below on the lowest-scoring pillar only. Ask 3-4 targeted questions. Update the scorecard if the deep-dive changes your assessment.

If they want the survey: Produce the Team Survey Template below.

If they want to move on: Route to the next skill.

Step 6: Route to Next Skill

Based on the scorecard, recommend ONE next step. See the Next Skill section below.

Deep-Dive Probes

Use these ONLY if the leader opts for the deeper assessment on a specific pillar. Ask 3-4 questions from the relevant section.

Psychological Barriers Deep-Dive

• Who on your team actively avoids AI tools? What's their role and seniority?
• When they explain why they don't use them, what do they actually say? (Get exact words)
• Do seniors and juniors behave differently around AI tools?
• Has anyone — even jokingly — said something about AI replacing them?
• How do your managers talk about AI tools? Do they use them themselves?

What you're listening for:

Signal	Barrier Type	Severity
"It writes bad code" (from senior who hasn't tried it)	Identity threat	High
Engineers quietly ignoring tools	Passive resistance	Medium
"Is this going to replace us?"	Fear of replacement	High
Juniors use it, seniors don't	Seniority-identity split	High
Managers mandate it without using it themselves	Leadership credibility gap	High

Integration Deep-Dive

• Walk me through a specific time someone tried an AI tool and stopped. What happened?
• How long did they try before deciding it didn't work?
• Did anyone configure the tool for your specific codebase, or was it out-of-the-box?
• When results are inconsistent, can people explain what changed between good and bad output?
• Is there a task your team hates doing that happens every sprint?

What you're listening for:

Signal	Issue	Severity
"Suggestions are wrong for our codebase"	Context/config problem	High — fixable
"I only use it for boilerplate"	Narrow use case adoption	Medium
"Results are inconsistent"	Variation confusion — people quit when they can't explain why	High
"Too many steps to use it"	Friction problem	Medium

Ownership Deep-Dive

• When something goes wrong with AI tool adoption, who hears about it?
• If I asked you right now how many people used an AI tool this week, could you tell me?
• Has anyone set a goal like "X people using Y tool by Z date"?
• Who would you call the AI adoption champion on your team?

What you're listening for:

Signal	Gap	Severity
"Me, plus my 5 other jobs"	No dedicated owner	Critical
"We have license count" as the only metric	No real measurement	Critical
No feedback channel for tool issues	No feedback loop	High
"We'll figure it out organically"	Wishful thinking	Critical

Anti-Patterns

"We're Pretty Early"

Symptom: Leader gives vague self-assessment to avoid specifics. Consequence: Assessment stays surface-level, scorecard is useless. Fix: "Early is fine — let me understand what 'early' looks like specifically. Of your team, how many have tried any AI tool even once?"

"Just Tell Me What to Do"

Symptom: Leader tries to skip assessment and get straight to solutions. Consequence: Generic advice that doesn't fit their actual situation. Fix: "This is 9 quick questions — under 5 minutes. It makes everything after this specific to your situation instead of generic advice."

"Our CTO Will Handle It"

Symptom: Leader delegates AI adoption without defining what "handle it" means. Consequence: Ownership gap goes undiagnosed because someone's name is attached. Fix: "What specifically are they doing on this? What's their timeline? What metrics are they tracking?" If the answer is vague, it's still an ownership gap.

Advice Creep

Symptom: You start suggesting tools, processes, or plans mid-assessment. Consequence: Diagnosis gets contaminated. Leader anchors on premature recommendations. Fix: STOP. Return to the next question. Recommendations come from other skills.

Output

After all questions, produce the scorecard in this exact format:

## AI Fluency Scorecard
**Company:** [name] | **Team:** [size] | **Date:** [date]
**Team stability:** [Stable / Restructuring in progress — [description] / Restructuring completed [timeframe] — [description]]

### Scores (1-5 scale)

| Pillar | Score | One-line summary |
|--------|:-----:|------------------|
| Psychological Barriers | X/5 | [key finding] |
| Integration Failures | X/5 | [key finding] |
| Ownership Gaps | X/5 | [key finding] |
| **Overall** | **X/5** | **[overall status]** |

### Fluency Levels
- **1** — No awareness or action. Team has access but hasn't engaged.
- **2** — Trying it out. Some individuals experimenting, but no consistency. People are still deciding if it's worth their time.
- **3** — Assisted use. A group uses AI regularly for specific tasks, but it's optional and individual. Effort sometimes feels disproportionate to results.
- **4** — Integrated. AI is embedded in team workflows, not just individual habits. Defaults have changed — some work no longer starts from scratch.
- **5** — Fluent. AI is invisible and expected. New hires pick it up by watching how the team works. Decisions are better, not just faster.

### Data Source
**Based on:** [Leader quiz / Imported data + gap-fill / Leader quiz + deep-dive / Leader quiz + team survey]
[If quiz-only: "These scores reflect your perspective. A 5-minute team survey would sharpen the picture — especially on psychological barriers, where leaders often underestimate resistance."]
[If imported data: "These scores combine your existing survey data with targeted questions to fill gaps. Data source: [describe what they shared]."]

### Top Finding
[Single most important thing this leader needs to address, in 2-3 sentences]

### Biggest Risk
[What will go wrong if they do nothing, in 1-2 sentences]

Next Skill

Based on the scorecard, recommend ONE next step:

Lowest-scoring pillar	Recommended next skill	Why
Psychological Barriers	blocker-diagnosis	Need to understand the specific resistance before acting
Integration Failures	first-use-case-picker	Tools aren't fitting — find the right starting point
Ownership Gaps	90-day-plan-builder	Need structure, owner, and metrics before anything else
Tied or all low	blocker-diagnosis	Start with people — tools and process follow

"Your biggest gap is [pillar]. I'd recommend running [skill name] next to [one sentence on what it does]. Want to do that now?"

Team Survey Template

If the leader wants the team survey, produce it in this format. This is sent to the team as a parallel track — it does NOT block the playbook from continuing.

## AI Tools — Quick Team Check (anonymous, 5 min)

1. How would you describe your current relationship with AI tools?
   - I haven't really started yet
   - I use them from time to time when someone points me to them
   - They're becoming a regular part of how I work
   - I'd feel less productive without them at this point

2. When you use AI, what does a typical interaction look like?
   - I ask a simple question and take the answer as-is
   - I give some background or instructions to shape the output
   - I go back and forth, adjusting my input until I get what I need
   - I work with templates, system prompts, or pre-built assistants for recurring tasks

3. What have you used AI for? (check all that apply)
   - I haven't used it for work yet
   - Writing, editing, or rephrasing text
   - Emails, messages, or communications
   - Looking things up or getting explanations
   - Brainstorming or structuring ideas
   - Working with data, spreadsheets, or reports
   - Preparing presentations or documents
   - Using an assistant or agent someone else set up
   - Building your own assistant, agent, or workflow
   - Automating repetitive steps in my workflow
   - Other: ___

4. How do you feel about AI becoming a bigger part of your work?
   - Skeptical — I'm not convinced it adds value for what I do
   - Curious — I'd like to understand what it can do before committing
   - Optimistic — I see real potential and want to go deeper
   - All in — I'm already pushing boundaries and exploring new use cases

5. What's the biggest reason you DON'T use AI tools more? (pick one)
   - I don't know what to use them for
   - The output isn't good enough for my work
   - It takes more effort than doing it myself
   - I don't trust the output
   - I haven't been given time to learn
   - I use them plenty already
   - Other: ___

6. Does your team have a shared way of using AI tools, or is it individual?
   - We have shared practices and standards
   - Everyone does their own thing
   - We don't really use them as a team

7. Who would you go to if you had a question or problem with an AI tool?
   - A specific person on the team
   - My manager
   - I'd figure it out myself
   - I wouldn't know who to ask

8. Is there a task in your daily work where you think AI could help but you're not sure how?
   [open text]

Mapping survey results to pillar scores:

Survey questions	Maps to pillar
Q4 (feelings), Q5 (reasons for not using)	Psychological Barriers
Q1 (relationship), Q2 (interaction depth), Q3 (breadth), Q5 (effort/output)	Integration Failures
Q6 (shared practices), Q7 (who to ask)	Ownership Gaps

References

• blocker-diagnosis — chains from this skill when psychological barriers are the main issue
• first-use-case-picker — chains from this skill when integration is the main issue
• 90-day-plan-builder — chains from this skill when ownership is the main issue
• full-adoption-cycle — orchestrates this skill as the first step in the complete sequence
• quarterly-review — re-runs this skill and compares to previous scorecard