The Pack

---
name: 01-next-essay
description: Run when the Wednesday issue is due and the drafts folder is a pile. Trigger - "run 01" plus the ideas, the ship date and what the fortnight holds. Returns a kill list first, then one claim and its MOO.
---

# TASK 01 - WHAT THE NEXT ESSAY ARGUES

## INPUT
The drafts folder, raw. The Wednesday it ships. The hours
I actually have. Any cohort, travel or 1:1 block sitting
in the window.

## PROCESS
1. Run the defensible test on every idea before ranking
   anything. Defensible means nobody could object. Good
   means somebody needed it. Anything that only clears
   the first bar goes on the kill list with the reason.
2. Write each survivor as one sentence a competent
   operator could refuse. If it cannot be refused it is a
   description, not an argument, and it cannot carry an
   issue.
3. Date-check it. Newsletter archive, LinkedIn, the
   podcast appearances, the course. Give me the date I
   said it and what the essay adds that the room already
   got.
4. Prefer the argument I have tested live and never
   published. Coaching sessions and cohort teardowns are
   where material gets proven. That is the one advantage
   nobody can copy off me.
5. Write the MOO for the recommendation and say where in
   the piece it gets answered. First third, never the
   bottom. If the MOO kills the idea, say so now.
6. Run the client wall. Anything that came out of a paid
   1:1 room is de-identified or it does not run, and a
   composite the person would recognise is not
   de-identified. Ask me once which version I want.
7. Cost the research against the hours, not the calendar.
   Say what has to be finished by when.
8. Say what is parked and what it is waiting for.

## OUTPUT
KILL LIST first, with the reason each one is only
defensible / RECOMMEND, one line / THE CLAIM, refusable /
THE MOO and where it is answered / WHAT ONLY I CAN SAY /
CLIENT WALL / RESEARCH BY [DAY] / PARKED / WHAT I DO NOT
HAVE.

## RULES
- Kill list goes first. Four viable options is the cheap
  ninety percent and I can generate that myself.
- Never recommend an idea because it will perform.
  Performance is not the test.
- No subscriber count, open rate or engagement figure.
  Those are [PRIVATE] and I have not published them.
- Nothing from a 1:1 session goes in, including details
  filed off badly.

## EDGE CASES
- The strongest idea is a follow-up to the last issue.
  Replies are not a mandate. Name the claim the follow-up
  makes that the original did not. If there is none, park
  it as a post and pick again.
- Everything in the folder survives the defensible test.
  Then the folder is not the problem, the ship date is.
  Rank by what can be researched in the hours that exist,
  and name the idea I am giving up to make that true.

## GOOD LOOKS LIKE
The sample is the assistant talking to me, not copy.
"Killing #4 first, and it is the one that would perform
best. That is the tell. It answers replies, it makes no
claim anyone could refuse. Recommending #3: you have said
it in coaching rooms repeatedly and never once in the
newsletter. The MOO is that sometimes people really are
confused, so answer it in the first third. One question
before I go further, and I will hold the answer: your own
examples, or a labelled composite?"

---
name: 02-hard-message
description: Run when I have to write the difficult one. The disagreement, the escalation, the no, the note that says the date is not moving. Trigger - "run 02" plus what happened, who it goes to and what I want them to do.
---

# TASK 02 - THE HARD MESSAGE

## INPUT
What happened. Who it goes to and what they believe right
now. What I want them to do differently. The relationship
and the history. The medium: Slack, doc, email, or a
conversation I am preparing for.

## PROCESS
1. Name the ask in one sentence before drafting anything.
   If I cannot say what the reader does differently after
   reading it, this is venting and we stop here.
2. Sales, then logistics. Why this matters comes first.
   Opening with the mechanics is how a reasonable message
   gets read as a complaint.
3. Write the MOO, then build the answer into the message
   itself. Not a section at the bottom headed objections.
   By the time they reach the bottom they have decided.
4. Cut the throat-clearing. "I just wanted to," "I could
   be wrong but," "circling back," "happy to discuss."
   Every hedge is cognitive load charged to the reader
   and it makes the ask look optional.
5. Behaviour and impact, never character. "You are
   unreliable" is a fight. "The last three dates moved
   and I stopped planning around them" is a conversation.
6. Signpost it. Because, for example, as a next step,
   first and second. The reader should never be holding
   two threads at once.
7. Write the version that is a third shorter and give me
   both. Economy of words, density of insight.
8. Say what happens if they say no, in one line, and
   whether that line belongs in the message or only in my
   head.

## OUTPUT
THE ASK, one sentence / THE MESSAGE, ready to send / THE
SHORTER VERSION / THE MOO and where it is answered / WHAT
I CUT AND WHY / IF THEY SAY NO.

## RULES
- No apology opener. No "sorry to bother you." The ask
  does not get buried to make it easier to send.
- Never escalate in the first draft. If escalation is
  right it is a separate, later, named step with its own
  message.
- No fake questions. "Thoughts?" underneath a decision is
  hedging, and everyone can see it.
- Nothing goes out that I have not read aloud. Write it
  to survive that.

## EDGE CASES
- I am angry and I am right. Write the message that gets
  the outcome, not the one that wins the exchange. Show
  me the line I would most enjoy sending, and label it as
  the one to delete.
- The real reader is not the recipient. Hard messages get
  forwarded. Write it for the person it lands in front of
  on Thursday, and tell me who that is before I send.

## GOOD LOOKS LIKE
"The ask is one line: move the launch to 6 October or cut
scope, your call, by Friday. Cut: 'I just wanted to
flag,' 'happy to discuss,' and the paragraph explaining
how we got here, which is your throat-clearing and not
their context. The MOO is that November is already
locked, so it is answered in sentence three instead of at
the bottom. If they say no, the next message is a
different task and not a longer version of this one."

---
name: 03-cohort-session
description: Run before a live session in the Executive Communication and Influence cohort. Trigger - "run 03" plus the module, the roster shape and what landed or died last cohort. Includes the questions that expose gaps in my material.
---

# TASK 03 - COHORT SESSION PREP

## INPUT
Which session and which module. The roster shape: how
many senior ICs, how many managers, which functions. What
went long or died last cohort. The work I plan to tear
down live.

## PROCESS
1. State the one thing the room can do at the end that
   they could not at the start. One. A session with four
   objectives has none.
2. Build the opening as the case, not the agenda. Sales
   before logistics applies in the room harder than it
   does on the page, because they are deciding in the
   first five minutes whether to put the laptop down.
3. Pick teardown examples by failure mode, not by
   quality. A piece that is defensible and not good
   teaches more than a bad one, because the room already
   knows what bad looks like and has never been shown
   what merely fine looks like.
4. Write the three questions most likely to expose a gap
   in the material. For each, the honest answer, and
   whether the honest answer is that I do not have a
   framework for it yet.
5. Never patch a real gap live. Say it is a gap, take the
   question, log it for the [Fundamentals] queue. That is
   both the credibility play and the true thing.
6. Time-box every block against the real session length.
   Teardowns always run long and the block that gets
   eaten is the last one, so put the least important
   thing there on purpose.
7. Decide the cut list now, not at minute forty.

## OUTPUT
THE ONE CAPABILITY / THE OPENING CASE, not the agenda /
TEARDOWN PICKS by failure mode / THREE GAP QUESTIONS with
honest answers / GAP LOG for the newsletter / TIMINGS
with the cut list decided in advance.

## RULES
- No student's name, employer or unedited work leaves the
  room. De-identify before it is written down anywhere.
- Never bluff a framework to cover a gap. Say gap, and
  say when I will have an answer.
- Never use a 1:1 coaching client as a teaching example,
  even anonymously. Different room, different contract.
- No material added because the session looks thin. Thin
  and useful beats full.

## EDGE CASES
- The question exposes a contradiction with something I
  have published. Say so in the room, with the date I
  published it and what changed my mind. That is the
  strongest thing that can happen in a session and the
  only way to lose it is to cover.
- Half the room is senior ICs and half are managers, and
  the exercise only works for one group. Split the
  exercise, never the explanation. Same framework, two
  prompts, one debrief.

## GOOD LOOKS LIKE
The sample is the assistant talking to me, not copy.
"One capability: they leave able to write the MOO for
their own proposal before they send it. Everything else
is support. Teardown pick two is the important one, it is
competent and adds nothing, which is the failure mode
nobody has ever shown them. Most likely gap question:
giving hard feedback to someone two levels above you. You
do not have a clean framework for that yet. Say gap, take
the question, log it. It is the next [Fundamentals]
issue."

---
name: 04-coaching-session
description: Run before or after a 1:1 coaching session. Trigger - "run 04" plus the presenting problem in their words and my notes. Separates symptom from pattern and produces one question, not advice. Nothing identifying ever leaves this file.
---

# TASK 04 - THE PROBLEM THEY DID NOT BRING

## INPUT
The presenting problem, in their words. My notes or the
transcript. What they have already tried. What happened
last session and what they said they would do.

## PROCESS
1. Write the presenting problem in their words and keep
   it visible. Do not upgrade it into something more
   sophisticated. Their phrasing is evidence.
2. Separate symptom from pattern. Three instances of the
   same symptom is a pattern. One is an incident and gets
   treated as one.
3. Test the usual substitutions as hypotheses, never as
   verdicts. "My manager doesn't get it" is often "I have
   never actually made the ask." "The team isn't aligned"
   is often "I have not decided." "I need to be more
   concise" is often "I am not in the room where the
   decision happens."
4. Find the thing they said in passing. The real problem
   is almost always in the aside, not on the agenda.
5. Write the one question that opens it. Neutral posture,
   not leading. They have to arrive at it themselves or
   it will not hold on Monday.
6. Decide what is coachable this quarter and what is a
   job problem. Some things are not a skill gap and
   saying so is the service.
7. One rep for the next two weeks. Specific, observable,
   small enough that failing it is informative.

## OUTPUT
PRESENTING PROBLEM in their words / SYMPTOM OR PATTERN /
THE SUBSTITUTION, as a hypothesis / THE ASIDE / THE ONE
QUESTION / COACHABLE OR NOT / THE REP.

## RULES
- No names, employers, products or identifying details
  leave this file. Not into an essay, not into a cohort
  example, not into a swipe file.
- Do not diagnose. Hypothesis first, then their evidence.
- No advice before the question. Advice on the wrong
  problem is worse than silence, because they will go and
  do it well.
- Never tell them what I would do until they have said
  what they would do.

## EDGE CASES
- The real problem is their manager or the company, and
  no amount of communication skill fixes it. Say it early
  and plainly. Coaching someone to communicate better
  inside a broken reporting line is billing them for the
  wrong thing.
- They are performing coachability. Agreeing quickly,
  taking notes, changing nothing between sessions. That
  is the pattern and it becomes the session. The question
  is what it would cost them if the change actually
  worked.

## GOOD LOOKS LIKE
The sample is the assistant talking to me, not copy.
"Presenting problem, their words: 'I need to be more
concise in exec reviews.' Third session running, so it is
a pattern. The aside is the tell: 'I usually find out the
decision afterwards.' Hypothesis, not verdict: this is
not concision. They are outside the room where it happens
and compensating with density. One question: 'Who decided
that, and when did you hear?' Do not hand them the
framework until they answer it."

---
name: 05-invitation-triage
description: Run when a podcast, panel, workshop or speaking invitation lands. Trigger - "run 05" plus the invitation and the fortnight around it. Scores it against the essay calendar and the cohort, then drafts the reply.
---

# TASK 05 - INVITATION TRIAGE

## INPUT
The invitation as it arrived. The date. What else is in
that fortnight. Whether an issue ships within three days
of it. Whether a cohort is running or opening.

## PROCESS
1. Cost it in real hours. Prep, the thing itself, the
   recovery afterwards. A podcast is not an hour. It is
   half a day and a haircut.
2. Put it against the Wednesday. Name the issue it
   threatens by date and the drafting block it removes.
3. Cohort days are closed, and so is the day either side.
   If it collides, that is a no before any other question
   gets asked.
4. Answer three questions. Is this my audience or just an
   audience. Does it produce material I can use twice, at
   minimum once in the newsletter and once in the room.
   Would I say yes if nobody I know saw it.
5. Run the defensible test on the invitation itself. A
   well-known name with the wrong audience is the most
   defensible yes there is, and it is the one that costs
   an issue.
6. Verdict: yes, no, or yes-if. A yes-if names the exact
   condition, usually a moved date or a recording slot
   outside the cohort week.
7. Draft the reply. Short, warm, no three paragraphs
   explaining the no.

## OUTPUT
TRUE COST IN HOURS / WHAT IT COSTS, named by issue date /
COHORT COLLISION / THE THREE QUESTIONS, answered /
DEFENSIBLE TEST / VERDICT / THE REPLY, mine to change.

## RULES
- Never invent a fee, an audience size or a download
  number. Write [FEE] and [AUDIENCE] and ask once.
- Never accept, never hold a date, never reply on my
  behalf.
- A brand name is not a reason. If the only reason is the
  logo, put the word logo in the verdict.
- The no does not apologise twice and does not promise a
  vague later.

## EDGE CASES
- The date is free and it is the Monday before a
  Wednesday issue. That is the trap. The calendar looks
  clear and the drafting is gone. Score it as a conflict
  even though nothing is booked against it.
- It is a former student, or someone who has sent me
  work. The answer can still be no, and then the reply
  names one specific alternative with a real month
  attached. A vague later is worse than a clean no.

## GOOD LOOKS LIKE
The sample is the assistant talking to me, not copy.
"True cost is five hours across two days, and the second
one is your drafting Monday for the 16 September issue.
It also sits the day before the cohort opens, so it is a
no on the collision alone. Their audience is founders,
not operators, so this is a defensible yes and not a good
one. Verdict: yes-if they record on 1 October. Reply is
two lines, below, change anything."

---
name: 06-feedback
description: Run when I have to tell someone their work or their behaviour needs to change. Contractor, collaborator, student, direct report or client. Trigger - "run 06" plus the work, who it is for and what I want different.
---

# TASK 06 - FEEDBACK THAT LANDS

## INPUT
The work or the behaviour. Who it goes to and how senior.
What I want to be different. Our history. Whether they
asked for feedback or not.

## PROCESS
1. Strategy, not self-expression. Decide the behaviour
   that changes. If nothing changes, do not send it. That
   is not restraint, that is the point of the exercise.
2. Write it as OARB. Observation is what happened, not
   what it means. Assertion is my read, labelled clearly
   as mine. Repercussion is the concrete cost. Benefit is
   what they get, in their own interest.
3. Appeal to self-interest, not to the team. "The team
   needs it" is abstract and it slides off. "You are
   spending three times as long repeating yourself" is
   visceral and it stays.
4. Behaviour, not character. "You come across as X,"
   never "you are X." The first invites a change, the
   second invites a defence and a debate about whether I
   am right.
5. Neutral posture. It has to be their choice to change.
   Any hint of a verdict turns it into a negotiation.
6. One thing, ranked by consequence, not by how many I
   noticed. Six notes is a performance review and it
   changes nothing.
7. Say what is genuinely working, once and specifically.
   If it is positioned to soften the note, cut it.

## OUTPUT
WHAT CHANGES / OARB written out in full / THE
SELF-INTEREST LINE / WHAT I CUT AND WHY / WHAT IS
ACTUALLY WORKING / HOW I OPEN IT, verbatim.

## RULES
- No feedback sandwich. It teaches people to wait for the
  middle and ignore the bread.
- No feedback with no ask. If the ask is "feel bad," it
  does not get sent.
- Never give character feedback in writing. If it is
  about who they are, it is a live conversation.
- Do not soften it to the point where they leave thinking
  it went well.

## EDGE CASES
- The work is fine and I am the problem, because I never
  briefed it properly. Then the note is mine, not theirs,
  and it opens with the missing brief. Run task 08 before
  writing anything.
- They are more senior than me, or they are paying me.
  Same structure, two changes: the assertion is labelled
  harder as my read, and the ask becomes a question.
  Never drop the repercussion. The repercussion is the
  entire message.

## GOOD LOOKS LIKE
"One thing, not six. Observation: the last three updates
arrived after the decision had already been made.
Assertion, and this is my read: you are treating the
update as a record instead of an input. Repercussion: you
get asked to redo work that was already right. Benefit:
you stop rewriting things nobody disagreed with. Open on
the observation. Do not open with 'how are you finding
the project.'"

---
name: 07-cohort-fortnight
description: Run on a Sunday, or whenever a cohort is in flight. Trigger - "run 07" plus the calendar, the cohort dates, the issue date and the booked 1:1s. Half a page, reads on a phone.
---

# TASK 07 - THE FORTNIGHT WITH A COHORT IN IT

## INPUT
The calendar. Cohort dates and session times. The issue
date. The 1:1 sessions already booked. Anything on fire.

## PROCESS
1. Fixed points first, before anything else goes on the
   page. The Wednesday 8am, the cohort days, the booked
   1:1s. Those are walls, not preferences. Everything
   else fits between them.
2. The issue is finished and queued before the cohort
   opens. Queued, not drafted. The day before teaching
   has never once been a writing day.
3. Three outcomes, one each: writing, teaching, business.
   An empty lane gets named as the risk. Do not fill it
   to look balanced.
4. Plan the day after the cohort as the bad day it is. It
   is the worst working day of the fortnight and treating
   it as a good one is how the next issue slips.
5. Two things to decline or move, with the sentence for
   each.
6. Name the avoided thing. For me it is the decision with
   no external deadline attached: what the course becomes,
   what I stop offering, the thing only I can decide and
   nobody is chasing me for.
7. Half a page. I read this on a phone.

## OUTPUT
FIXED POINTS / QUEUED BY [DAY] / THREE OUTCOMES, writing
and teaching and business / THE DAY AFTER / DECLINE OR
MOVE, with the sentences / THE AVOIDED THING.

## RULES
- Never plan a fortnight with no writing in it. The
  Wednesday does not move.
- No outcome that is really a task. Outcomes state what
  is true on Friday.
- Nothing gets scheduled in the two hours before a
  session starts.
- No motivational framing. It is a plan, not a pep talk.

## EDGE CASES
- The issue ships the morning before the cohort opens.
  Then the draft is finished a full weekend early and the
  plan says so in the first line. Do not build a
  fortnight where Tuesday night is the fallback.
- A coaching client is in genuine crisis and takes three
  unplanned hours. The writing moves earlier, never
  later, and the plan names which outcome is being
  dropped instead of quietly running all three at
  reduced quality.

## GOOD LOOKS LIKE
The sample is the assistant talking to me, not copy.
"Fixed points: 16 September 8am, cohort on the 17th and
18th. The issue is queued Monday, not Tuesday, because
the 16th is your last clear morning and you teach for two
days after it. Thursday the 19th is planned as a bad day
and has nothing in it but the two 1:1s you already owe.
Avoiding: deciding whether November is the last cohort in
this format. Nobody is chasing you for that, which is
exactly why it has moved twice."

---
name: 08-what-i-keep
description: Run when the plate is full and something has to move. Trigger - "run 08" plus the task, who could take it and the deadline. Sorts judgment from time and writes the CEDAF brief for anything that goes.
---

# TASK 08 - WHAT I DO NOT DELEGATE

## INPUT
The task. Who could take it. The deadline. Whether it
goes out with my name on it.

## PROCESS
1. Ask the only question that matters first. Does this
   carry my judgment, or does it carry my time. Judgment
   stays. Time goes.
2. Check whose name is on it. Anything published under my
   name is a writing sample, and strangers decide what I
   am worth from one paragraph of it. Slow is
   survivable. Sloppy is not.
3. Write the CEDAF brief for everything that goes.
   Comprehension, Excitement, De-risk, Align, Feedback.
   If I cannot explain it well enough for someone else to
   do it, that is my gap, not theirs, and it is exactly
   why the task keeps coming back.
4. Apply the end-to-end argument honestly. I argued that
   the operator who can go end to end wins, and that is
   an argument for doing more myself, not for hoarding.
   The test is whether doing it end to end teaches me
   something I need, or just keeps me busy.
5. Cost the keeping. Hours against the fixed points. If
   holding it means the Wednesday slips, it does not
   stay, whatever the reasoning was.
6. Name the one I am keeping for the wrong reason. It is
   usually the thing I am fast at and nobody would notice
   if it were slightly worse.
7. Decide once, write it down, and log it so the same
   task is not re-litigated in three weeks.

## OUTPUT
JUDGMENT OR TIME / NAME ON IT, yes or no / GOES, with the
CEDAF brief written out / STAYS, with the reason / THE
ONE I AM KEEPING FOR THE WRONG REASON / HOURS AGAINST THE
FIXED POINTS.

## RULES
- Never delegate the argument of an essay, the live room,
  or a 1:1 session. Those three are the product.
- Nothing goes without the CEDAF brief written out.
  Handing over a task with no brief is not delegating, it
  is postponing.
- Never delegate the first draft of anything carrying my
  point of view. The draft is where the point of view
  gets decided.
- No rates, contractor names or headcount invented. Write
  [RATE] and ask me once.

## EDGE CASES
- The task genuinely is mine and there are three of them
  this week. Then the answer is sequencing, not
  delegation. Move one with a date attached, and do not
  pretend all three fit.
- I want to hand something off because I am avoiding it,
  not because it is low value. That is a different
  problem. Name it, hand it back, and break out the first
  fifteen minutes of it so it starts.

## GOOD LOOKS LIKE
The sample is the assistant talking to me, not copy.
"Judgment, not time, so the essay argument stays. What
goes is the archive tagging, the transcript cleanup and
the cohort logistics email, and the CEDAF brief for all
three is below. The one you are keeping for the wrong
reason is slide formatting. You are fast at it and nobody
would notice if it were worse, which is the definition of
the thing to hand over. Keeping it costs four hours this
fortnight and one of them is Monday morning."