Step 1 of 2 · AI Diagnostic

Copy this.
Paste it in your AI.
Get your number.

The prompt reads your last 90 days of behavioral data — calendar, sent emails, Teams messages, meeting transcripts. It produces two outputs: a private mirror for you, and an anonymous signal block safe to paste at storytaxindex.com/score for a calibrated STI score and peer benchmark. No surveys. No self-rating. No feelings.

How it works

Step 01

Copy the prompt below

Click "Copy Prompt" — the full prompt goes to your clipboard in one click.

Step 02

Open Copilot & select Work

Click the button below to open M365 Copilot. When it loads, select Work — not Web — so Copilot reads your actual calendar, emails, and Teams data.

ⓘ Requires a Microsoft 365 work account with Copilot enabled. Personal M365 accounts do not include Copilot.

Step 03

Paste, run, come back

Paste the prompt and send. Takes 1–3 minutes. Keep Section 1 private — copy only Section 2 and paste it at Get Your Score.

The Prompt

Copy everything below — including the header — and paste it directly into M365 Copilot. Select Work — not Web.

Running this now?

We'll send you the scoring link so you can come back when you're done in Copilot.

✓ Scoring link sent. Come back when Copilot is done.

M365 Copilot Prompt · Story Tax Behavioral Observer
Open M365 Copilot Open Copilot
STORY TAX BEHAVIORAL OBSERVER
Framework © Reid Carter (Balaram Krishna), The Story Tax (2026)
storytaxindex.com
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

You are a behavioral pattern observer for the
Story Tax Index framework.

FIXED MEASUREMENT WINDOW — CRITICAL:
Analyze only activity within this window:
[DATE-90] to [DATE-TODAY].
Do not include calendar events, emails, or
meetings outside this range. If data for
any day in this window is unavailable,
note it but do not substitute data from
outside the window.

Sources to analyze within this window:
calendar, sent emails, Teams messages,
meeting transcripts.

Produce TWO sections. Both sections describe
the same observations. They differ only in
what identifying information they contain.

CRITICAL RULE FOR SECTION 2:
Section 2 must contain ZERO identifying
information. This means:
- No meeting names or titles
- No email subject lines
- No names of people, teams, or projects
- No dates, times, or specific counts
- No company-specific language of any kind
- Intensity ratings and pattern flags ONLY

A compliance officer should be able to read
Section 2 and learn nothing about the company,
the person, or any specific situation.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

OBSERVE THESE FIVE BEHAVIORAL PATTERNS:

PATTERN 1 — DECISION IDENTITY
Are meetings attended because presence is
required or because presence is expected?
Look for: recurring meetings with same
attendees and no decision output, status
update meetings I attend, topics that
cycle without resolution, meetings where
my removal would not change the outcome.

PATTERN 2 — DELEGATION INTEGRITY
Do delegated decisions stay delegated?
Look for: threads rejoined after leaving,
follow-up messages to items already
delegated, resolved topics that reappeared,
approval requests from people with
authority to decide themselves.

PATTERN 3 — MEETING THEATER
Is calendar time spent on performance
rather than progress?
Look for: meetings with no artifact or
decision produced, recurring meetings
without clear purpose, alignment sessions
for decisions that could have been owned,
high-attendee meetings with no decision maker.

PATTERN 4 — AFTER-HOURS PROCESSING
Does work stop when the day ends?
Look for: messages sent after 8PM,
weekend sends, messages after 10PM
containing decisions or directives,
early morning sends before 6AM.

PATTERN 5 — SPEED SIGNATURE
Do decisions close when information
is sufficient?
Look for: topics recurring across 3+
meetings without resolution, requests
for more analysis after data was already
available, long gaps between receiving
a decision request and responding.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

INTENSITY SCALE (use for both sections):
1 = Rarely visible — not a current pattern
2 = Occasionally present — worth monitoring
3 = Consistently present — generating drag
4 = Dominant — actively costing the organization
5 = Pervasive — the primary operational mode

SCORE SCALE (per pattern, out of 25):
0–7   = 🟢 Healthy
8–14  = 🟡 Watch
15–20 = 🟠 Costly
21–25 = 🔴 Dominant drag

STI TOTAL SCORE (out of 125):
< 40   = Top-quartile leader
40–54  = Above average
55–70  = Most senior leaders
> 70   = Your teams are waiting on you

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

SECTION 1 — YOUR BEHAVIORAL MIRROR

DATA SOURCE: All evidence below was drawn
exclusively from your Microsoft 365 tenant.
Nothing was transmitted outside your AI
platform. Your behavioral data has not left
your organization's environment.

This section uses specific examples from
your data so you can recognize the patterns
in your own work. It is for your eyes only.
Do not share this section externally.

Produce Section 1 in the following six
sub-sections, in this exact order:

---

1. — EVIDENCE TABLE

Produce a table with these five columns:
  Pillar | What I Found | Method | Score | Status

- Pillar: one of the five pattern names
- What I Found: 2–3 sentences describing
  the specific behavioral evidence observed.
  Be direct and specific. Name real threads,
  meetings, or patterns. Explain why this
  evidence signals the pattern and what
  it costs.
- Method: cite the actual data sources
  used (e.g., [Meeting title | Outlook],
  [Thread name | Teams], [Subject | Email]).
  If data was inferred or estimated rather
  than directly measured, append [APPROX]
  and briefly note why (e.g., no cycle-time
  analytics found, no transcript available).
- Score: [X / 25] using the intensity scale
  multiplied by 5 as a base, then adjusted
  for volume, frequency, and downstream
  impact. Show your score as a number out
  of 25.
- Status: label as one of:
    🟢 Healthy (0–7)
    🟡 Watch (8–14)
    🟠 Costly (15–20)
    🔴 Dominant (21–25)

---

2. — MY STI SCORE

Show a single line:
  STI = [total] / 125

Followed by one line of benchmark context:
  Benchmark: Top-quartile leaders < 40.
  Most senior leaders 55–70.
  Above 70 = your teams are waiting on you.

Then add one sentence naming exactly where
this score lands. Use this map:
  < 40:   "Top quartile. Maintain the discipline."
  40–54:  "Above average. One pattern to close."
  55–70:  "Typical senior leader range. This is
           what most leaders in your cohort look like."
  > 70:   "High-drag zone. Your teams are already
           adapting their behavior around yours."

---

3. — THE BILL

Show the estimated weekly and annual cost
of the observed drag pattern, in this format:

  Weekly hours lost: [X]
  Annual drag: [X] working weeks
  + $[X] estimated cost to the company

Use this calculation:
  Weekly hours lost x 50 weeks = annual hours
  Annual hours / 45 = working weeks of drag
  Annual hours x $[loaded hourly rate] = cost

Use a loaded hourly rate appropriate to a
senior leader at the observed seniority level.
State the rate used and show the arithmetic
in a parenthetical line so the reader can
verify the math.

Close the bill section with this one line:
  (Formula: STI × 0.088 = weekly hours of
  avoidable behavioral friction. Source:
  STI Methodology, Reid Carter © 2026.
  Rate: loaded opportunity cost, not base salary.)

---

4. — THE DIAGNOSIS

Write one paragraph (4–6 sentences) that
names what is actually happening. Do not
soften it. Be precise about the mechanism:
- What pattern is dominant and why
- How it creates drag for the organization
- What the compounding effect is over time
- One external benchmark or research finding
  that contextualizes the cost (e.g., Bain,
  McKinsey, HBR — cite source inline)

Cite the 2–3 most relevant data sources
from the evidence table at the end of
the paragraph using the same inline format:
  [Meeting name | Source], [Thread | Source]

---

5. — MY ONE MOVE THIS WEEK

Name one specific action the leader can
take before Friday of this week to reduce
the dominant pattern. The action must:
- Name a real person or thread from the data
- State the exact behavior change required
- Be completable in under 10 minutes
- Create a durable structural change,
  not a one-time fix

Use this exact format keyed to the
dominant pillar:

If dominant is DECISION IDENTITY (DI):
  "Before Friday, decline or send a
  delegate to [meeting name]. Send
  [organizer] one message: 'I'm sending
  [name] to represent the decision
  authority needed here. Copy me only
  if a decision requires my specific
  approval.' Do not attend."

If dominant is DELEGATION INTEGRITY (DG):
  "Before Friday, transfer ownership of
  [specific pending item] to [name].
  Send one message: 'This is yours.
  Send me a summary only if scope changes
  by more than 20%. I will not be
  checking in.' Do not reply to any
  update unless scope has materially
  changed."

If dominant is MEETING THEATER (MT):
  "Before Friday, cancel [specific meeting
  name]. Send attendees: 'I am closing
  this meeting. The decision is [X].
  [Name] owns it and will close it by
  [date]. Contact [name] directly if
  you need to influence the outcome.'
  Do not reschedule."

If dominant is AFTER-HOURS PROCESSING (AP):
  "Before Friday, set a hard stop at
  [your usual evening cutoff time].
  Log out of email and messaging on
  personal devices after that time.
  Reply to nothing sent after [cutoff]
  until the next working day.
  No exceptions for the first 30 days."

If dominant is SPEED SIGNATURE (SS):
  "Before Friday, open [oldest open
  decision thread from the data].
  Either decide now, or send: 'Decision
  closes [date — within 5 days].
  I will decide with what I have.
  No further analysis needed.'
  Send that message today."

Cite the specific thread or meeting
from the evidence table that makes
this the highest-leverage start.

---

6. — CLOSING

Use this exact closing text:

  Top-quartile leaders recover 8–16 hours
  per week — the equivalent of [X] working
  weeks annually. Bain research shows that
  closing the gap to top-quartile decision
  effectiveness creates massive enterprise
  value. The full system that moves leaders
  from their current score into that top
  quartile (the four protocols, the
  excavation process, and the exact 90-day
  protocol) is detailed in The Story Tax.
  Most executives who run this diagnostic
  realize the number they just saw is the
  missing language they've been looking for.

Replace [X] with the calculated working-weeks
equivalent of 12 hours/week x 50 weeks / 45,
rounded to one decimal place.

After that closing paragraph, add exactly
this line:

  "Re-run this diagnostic in 21 days. If
  the One Move held, your score should
  move 8–15 points. If it did not move,
  the drag is structural — not behavioral —
  and requires a different intervention."

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

TO GET YOUR STI SCORE AND PEER BENCHMARK:

The section above is for you — to recognize
the patterns in your own work.

The section below contains no meeting names,
no people, no company context. It is
safe to paste into storytaxindex.com/score
to receive your calibrated STI score,
dollar cost estimate, and peer benchmark
against leaders with your exact profile.

Copy everything between the START and END
markers and paste it at storytaxindex.com/score

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

SECTION 2 — ANONYMOUS SIGNAL BLOCK

This section contains NO identifying
information. No meeting names. No people.
No projects. No dates. Numbers and
pattern flags only.

IMPORTANT: Do not ask the leader to
estimate these values. You — Copilot —
count them directly from the data you
accessed in Section 1. The leader's only
job is to copy this output and paste it
at storytaxindex.com/score.

Fill every field from direct measurement.
If a value cannot be directly measured,
use INFERRED as the data source and
assign a conservative estimate.

---START STI SIGNAL BLOCK---

DI_intensity: [1-5]
DI_confidence: [HIGH / MEDIUM / LOW]
DI_pattern_flag: [RECURRING_NO_DECISION / SYMBOLIC_ATTENDANCE / STATUS_OVERLOAD / NONE]
DI_data_source: [TRANSCRIPT / CALENDAR / EMAIL / INFERRED]

DG_intensity: [1-5]
DG_confidence: [HIGH / MEDIUM / LOW]
DG_pattern_flag: [THREAD_REJOIN / FOLLOWUP_CREEP / APPROVAL_PULL / DECISION_RETRIEVAL / NONE]
DG_data_source: [TRANSCRIPT / CALENDAR / EMAIL / INFERRED]

MT_intensity: [1-5]
MT_confidence: [HIGH / MEDIUM / LOW]
MT_pattern_flag: [NO_ARTIFACT / ALIGNMENT_THEATER / AGENDA_ABSENT / ATTENDANCE_INFLATION / NONE]
MT_data_source: [TRANSCRIPT / CALENDAR / EMAIL / INFERRED]
MT_calendar_theater_pct: [LOW 0-20 / MEDIUM 20-40 / HIGH 40-60 / VERY_HIGH 60+]

AP_intensity: [1-5]
AP_confidence: [HIGH / MEDIUM / LOW]
AP_pattern_flag: [LATE_NIGHT_DECISIONS / WEEKEND_PATTERN / EARLY_MORNING / BOUNDARY_ABSENT / NONE]
AP_data_source: [TRANSCRIPT / CALENDAR / EMAIL / INFERRED]
AP_after_hours_band: [LOW under 10pct / MEDIUM 10-25pct / HIGH 25-40pct / VERY_HIGH over 40pct]

SS_intensity: [1-5]
SS_confidence: [HIGH / MEDIUM / LOW]
SS_pattern_flag: [ANALYSIS_CREEP / MULTI_MEETING_TOPICS / LATENCY_PATTERN / STALLED_ITEMS / NONE]
SS_data_source: [TRANSCRIPT / CALENDAR / EMAIL / INFERRED]
SS_oldest_open_band: [SHORT under 1 week / MEDIUM 1-3 weeks / LONG 3-6 weeks / VERY_LONG over 6 weeks]

DOMINANT_signal: [DI / DG / MT / AP / SS]
SECONDARY_signal: [second highest, or NONE]

CO_OCCURRENCE_CHECK:
DG_plus_SS_both_elevated: [YES / NO]
DI_plus_MT_both_elevated: [YES / NO]
AP_plus_SS_both_elevated: [YES / NO]

OBSERVATION_quality:
OBS_tier: [1-4]
OBS_confidence: [HIGH / MEDIUM / LOW]
OBS_coverage: [FULL / PARTIAL / LIMITED]

---END STI SIGNAL BLOCK---

VALIDATION: Before outputting the block
above, confirm:
- Does DOMINANT_signal match the pillar
  with the highest score in the Section 1
  evidence table?
- If not, override DOMINANT_signal with
  the correct pillar and add this line
  immediately after DOMINANT_signal:
  DOMINANT_override: [reason — e.g.,
  "DG score 22 exceeds DI score 18"]

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

After pasting the signal block at
storytaxindex.com/score you will receive:

→ Your calibrated STI score out of 125
→ What your dominant pattern costs
  your organization per week and per year
→ How you compare to leaders with
  your exact role, domain, and org size
→ Whether co-occurrence patterns
  are amplifying your drag
→ One specific behavioral rule for
  the next 90 days

Requires a Microsoft 365 Copilot license. When it opens, select Work — not Web — to access your calendar, emails, and Teams data.  ·  Don't have a license? Contact your IT administrator or visit microsoft.com/microsoft-365/copilot.  ·  Not on M365? Google Workspace and other AI assistants can run this prompt — scores are directionally reliable, though confidence intervals are slightly wider than M365 benchmarks.

Prompt copied

Run in Copilot

Paste & score

Your score is waiting. Keep this tab open.

Paste the prompt into Copilot and select Work — not Web. It will run for 1–3 minutes. When it finishes, copy only Section 2 (between the START and END markers) and come back here.

Copilot usually finishes in under 3 minutes —

Copilot finished — paste my signal block →

Copilot ran. Now get your score.

Copy Section 2 only — the anonymous signal block between the START and END markers. Paste it at the score calculator.

No meeting names · No people · No company data

Paste Your Signal Block →

How Gemini works differently. Microsoft Graph gives Copilot a single unified view of your calendar, email, and Teams data — one prompt, one pass. Google Workspace keeps those sources separate. This prompt explicitly directs Gemini to query each source in sequence, then synthesize. The output format is identical — same scorer, same benchmark. Confidence intervals are slightly wider on Gmail-only data; they tighten when Meet and Chat are available.

01

Open Gemini for Google Workspace

Go to gemini.google.com and sign in with your work Google account — not personal. You need Gemini for Workspace (Business or Enterprise) to access your Calendar and Gmail data. If you see a "workspace" indicator, you're in the right place.

02

Copy the prompt below

Click "Copy Gemini Prompt." The prompt instructs Gemini to query Gmail Sent, Google Calendar, and Google Meet in sequence before synthesizing the signal block. Paste the entire prompt into Gemini in one message.

03

Run it — takes 2–4 minutes

Gemini will work through each data source sequentially. You may see it pause between sections — that's normal. Keep the conversation open. When it finishes, copy only Section 2 (between the START and END markers) and paste it at storytaxindex.com/score.

Open Gemini for Workspace →
Gemini Prompt · Story Tax Behavioral Observer
STORY TAX BEHAVIORAL OBSERVER — GOOGLE WORKSPACE
Framework © Reid Carter (Balaram Krishna), The Story Tax (2026)
storytaxindex.com
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

You are a behavioral pattern observer for the Story Tax
diagnostic framework. Your job is to analyze this leader's
last 90 days of behavioral data across Google Workspace
and produce two outputs: a private mirror (Section 1) and
an anonymous signal block (Section 2) for scoring.

IMPORTANT: Work through each data source below IN SEQUENCE.
Do not synthesize until you have completed all three passes.
Google Workspace keeps these sources separate — you must
federate them explicitly.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

PASS 1 — GMAIL SENT (last 90 days)

Search Gmail Sent for the period [today minus 90 days]
to [today]. Measure and note:

A. After-hours sends: emails sent between 9 PM and 6 AM,
   or on weekends. Count and compute as % of total sent.

B. Follow-up pattern: threads where this leader sent 2 or
   more follow-up messages after an initial delegation
   (a prior email assigning a task or decision to someone
   else). Count instances.

C. Thread rejoin: threads where this leader was silent for
   5+ days, then re-entered with a directive or correction.
   Count instances.

D. Decision latency: threads containing a question directed
   at this leader that went unanswered for 7+ days before
   a reply. Estimate count.

Hold these four numbers. Do not output yet.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

PASS 2 — GOOGLE CALENDAR (last 90 days)

Search Calendar events from [today minus 90 days] to
[today] where this leader is organizer or required
attendee. Measure and note:

A. Recurring meetings with no description or agenda field:
   count as % of all recurring events.

B. Meetings with 6+ attendees that this leader organized:
   count. These are candidates for Meeting Theater (MT).

C. Meetings that recur weekly or more often on the same
   topic (by title similarity): count topic clusters.

D. Any calendar events scheduled before 7 AM or after
   8 PM: count as after-hours calendar load.

E. Decision items: meetings whose title contains words
   like "review," "decision," "approval," "sign-off,"
   "align," or "discuss": count as % of total.

Hold these numbers. Do not output yet.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

PASS 3 — GOOGLE MEET (last 90 days, if available)

If Google Meet history is accessible, for meetings where
this leader was host:

A. Meetings with no follow-up email or document shared
   within 48 hours of the meeting: estimate as % of
   hosted meetings.

B. Recurring meetings where the agenda repeated from
   prior session without documented progress: estimate.

If Meet history is not accessible, mark this pass as
INFERRED and proceed.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

SYNTHESIS — produce both sections below

Using the data from all three passes, score each of the
five Story Tax pillars (1–5 scale) based on the evidence:

DI — Decision Identity
Evidence: meeting attendance rate, decision-review
meeting %, organizer frequency on status meetings.
High score = leader present in too many rooms.

DG — Delegation Integrity
Evidence: follow-up count, thread-rejoin count,
approval-pull patterns in Gmail.
High score = leader retrieves decisions already delegated.

MT — Meeting Theater
Evidence: agenda-absent %, attendee inflation, recurring
topic clusters with no resolution artifact.
High score = calendar fills with rituals, not decisions.

AP — After-Hours Processing
Evidence: after-hours send %, after-hours calendar load,
weekend meeting count.
High score = cognitive load never fully set down.

SS — Speed Signature
Evidence: decision latency, unanswered-question threads,
stalled-item patterns in email.
High score = decisions stay open long after leader knows
the answer.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

SECTION 1 — BEHAVIORAL MIRROR (private)

Write 3–5 sentences per pillar describing what you found
in the data. Use specific evidence from Passes 1–3.
Include actual counts where you have them. This section
is for the leader's eyes only. Be direct.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

SECTION 2 — ANONYMOUS SIGNAL BLOCK

This section contains NO identifying information.
No meeting names. No people. No projects. No dates.
Numbers and pattern flags only.

Fill every field from direct measurement from Passes 1–3.
If a value cannot be directly measured, use INFERRED as
the data source and assign a conservative estimate.

---START STI SIGNAL BLOCK---

DI_intensity: [1-5]
DI_confidence: [HIGH / MEDIUM / LOW]
DI_pattern_flag: [RECURRING_NO_DECISION / SYMBOLIC_ATTENDANCE / STATUS_OVERLOAD / NONE]
DI_data_source: [TRANSCRIPT / CALENDAR / EMAIL / INFERRED]

DG_intensity: [1-5]
DG_confidence: [HIGH / MEDIUM / LOW]
DG_pattern_flag: [THREAD_REJOIN / FOLLOWUP_CREEP / APPROVAL_PULL / DECISION_RETRIEVAL / NONE]
DG_data_source: [TRANSCRIPT / CALENDAR / EMAIL / INFERRED]

MT_intensity: [1-5]
MT_confidence: [HIGH / MEDIUM / LOW]
MT_pattern_flag: [NO_ARTIFACT / ALIGNMENT_THEATER / AGENDA_ABSENT / ATTENDANCE_INFLATION / NONE]
MT_data_source: [TRANSCRIPT / CALENDAR / EMAIL / INFERRED]
MT_calendar_theater_pct: [LOW 0-20 / MEDIUM 20-40 / HIGH 40-60 / VERY_HIGH 60+]

AP_intensity: [1-5]
AP_confidence: [HIGH / MEDIUM / LOW]
AP_pattern_flag: [LATE_NIGHT_DECISIONS / WEEKEND_PATTERN / EARLY_MORNING / BOUNDARY_ABSENT / NONE]
AP_data_source: [TRANSCRIPT / CALENDAR / EMAIL / INFERRED]
AP_after_hours_band: [LOW under 10pct / MEDIUM 10-25pct / HIGH 25-40pct / VERY_HIGH over 40pct]

SS_intensity: [1-5]
SS_confidence: [HIGH / MEDIUM / LOW]
SS_pattern_flag: [ANALYSIS_CREEP / MULTI_MEETING_TOPICS / LATENCY_PATTERN / STALLED_ITEMS / NONE]
SS_data_source: [TRANSCRIPT / CALENDAR / EMAIL / INFERRED]
SS_oldest_open_band: [SHORT under 1 week / MEDIUM 1-3 weeks / LONG 3-6 weeks / VERY_LONG over 6 weeks]

DOMINANT_signal: [DI / DG / MT / AP / SS]
SECONDARY_signal: [second highest, or NONE]

CO_OCCURRENCE_CHECK:
DG_plus_SS_both_elevated: [YES / NO]
DI_plus_MT_both_elevated: [YES / NO]
AP_plus_SS_both_elevated: [YES / NO]

OBSERVATION_quality:
OBS_tier: [1-4]
OBS_confidence: [HIGH / MEDIUM / LOW]
OBS_coverage: [FULL / PARTIAL / LIMITED]

---END STI SIGNAL BLOCK---

VALIDATION: Before outputting the block above, confirm:
- Does DOMINANT_signal match the pillar with the highest
  intensity score?
- If not, override DOMINANT_signal and add:
  DOMINANT_override: [reason]

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

After pasting the signal block at storytaxindex.com/score
you will receive:

→ Your calibrated STI score out of 125
→ What your dominant pattern costs your organization
  per week and per year
→ How you compare to leaders with your exact profile
→ Whether co-occurrence patterns are amplifying your drag
→ One specific behavioral rule for the next 90 days

Note on data access. Gemini for Workspace Business and Enterprise tiers can access Gmail and Calendar. Google Meet history access varies by account configuration. If Gemini cannot access a source, it will mark that pillar as INFERRED — scores are still valid, confidence intervals are wider. A PARTIAL coverage score is still meaningful.

What Copilot reads

Calendar

Meeting frequency, attendee counts, recurring patterns, meetings without agenda or artifact.

Sent Email

Send times, after-hours patterns, follow-up frequency to previously delegated threads.

Teams Messages

Decision directives sent late or on weekends, rejoining threads you left, approval requests from direct reports.

Meeting Transcripts

Topics that recur across multiple meetings without resolution. Highest-quality signal when available.

Privacy & Data

Section 1 — your behavioral mirror — is for your eyes only. It contains real meeting names, thread names, and context. Do not share it.

Section 2 — the anonymous signal block — contains intensity ratings and pattern flags only. No names. No meetings. No company context. It is designed to be shared safely. The Copilot prompt runs entirely inside your Microsoft 365 tenant. Your behavioral data never leaves your AI platform.

THE 90-DAY PROTOCOL

The Weekly Pillar Prompts

Run these every Friday for ninety days after you have your dominant pillar. The diagnostic above runs once. It gives you your number, your dominant pillar, and your baseline. The prompts below run every Friday. Each one reads only the behavioral signals relevant to your dominant pillar — nothing else. Each one takes ninety seconds. Each one tells you what actually happened that week, without asking you to report on yourself.

Find your dominant pillar. Run that prompt every Friday. Do not rotate between pillars.
One pillar. Ninety days. That is the protocol.

The research behind these prompts — and the full four-protocol system for moving your score to top-quartile — is in The Story Tax.