Sales Runbook

Ask first

• Which sources may it read? (default your inbox, your texts, and notes from meetings if you keep them; ignore newsletters and automated mail)
• Who counts as someone you owe or who owes you? (default anyone who wrote to you personally in the last 60 days)
• When is a promise stale? (default 30 days, or 60 days if it has no date)
• Where should the list live? (default one page you open each morning)

Needs

• Read access to the sources
• One saved list with these fields for every item: title, who, direction (mine or theirs), source and link, date captured, due date if any, confidence (high or low), status (open, done or dropped) and why it closed
• A scheduler that can run every 15 minutes
• If your assistant cannot read one of these, paste it when it asks. A thinner version still works
• Any password or token is entered by you into your device's secure store, never pasted into a chat or saved in a plain file

Build

1. Every 15 minutes, read messages and notes added since the last run. Pick out sentences where someone commits to do something: "I will", "I'll send", "we'll have it by", "can you send me".
2. Save each as one item with every field above. Mark confidence low if the commitment is vague or the date is a guess.
3. When a later message in the same thread contains the promised thing (an attachment, a link, the words "here it is"), close the item and write why. Never close an item on silence.
4. Mark an item stale after the stale limit and move it to a dropped list, never delete it.
5. Show the list with days waiting, oldest first, amber after seven days. This one list is the source for every other run that needs promises: the morning brief reads its Due from you section from it, and the follow-up run adds each commitment it drafts to it.

Check

• Seed three promises in a test thread; confirm all three appear with the right direction and date
• Send the promised thing; confirm the item closes and says why
• Confirm a vague promise is marked low confidence and a stale one moves to the list of dropped items
• Say whether the writing step ran for real or on sample text. Check that every figure, quote and link matches its source, and drop any that does not

In my version

• Claude Code on a 15-minute schedule, one list I open each morning; my morning brief and my follow-up run both read and write it

Ask first

• Where are your past proposals?
• Where is the price list, and are there discounts you may offer? (default none)
• Who may see a draft? (default only you)

Needs

• Read access to notes, past proposals and the price list
• If your assistant cannot read one of these, paste it when it asks. A thinner version still works
• Any password or token is entered by you into your device's secure store, never pasted into a chat or saved in a plain file

Build

1. Every hour on weekdays, find discovery calls that ended in the last two hours. Keep a list of the items already handled (by their address or id) and skip any that are on it, so the same item never produces a second brief, comment or draft.
2. Pull the needs in their words.
3. Write the proposal in the style of your past ones using only listed prices.
4. List every question you could not answer. Never send it.

Check

• Run it on a call you remember; every price is on your list
• Confirm a draft was saved and nothing was sent
• Say whether the writing step ran for real or on sample text. Check that every figure, quote and link matches its source, and drop any that does not

In my version

• Claude Code on an hourly weekday schedule, a draft in my documents

Ask first

• Which accounts and contacts should it watch? (default anyone you emailed in the last 90 days with an open deal)
• How long is quiet? (default 10 working days)
• Where should it reach you? (default a message from your assistant)

Needs

• Any password or token is entered by you into your device's secure store, never pasted into a chat or saved in a plain file
• Read access to your inbox, calendar and an account list
• Permission to read public profiles; never to contact anyone
• If your assistant cannot read one of these, paste it into the chat when it asks. A thinner version still works

Build

1. Each morning, check each watched contact for a reply, a meeting, a job change and a public post.
2. List those quiet longer than the limit, and any job change or relevant post. Count a clear rise or fall in activity as a change of 25% or more against the 60-day average.
3. Add one suggested next step to each line, and leave it as a draft on that contact's existing thread. Never send it.
4. Send nothing on a day with no change. Save each flagged contact with the date, and do not flag the same contact for the same reason again within 14 days.

Check

• Run it against a contact you know changed jobs; confirm it is caught with a source link
• Confirm a quiet contact is listed once, not every day
• Say whether the writing step ran for real or on sample text. Check that every quote and link on the page matches its source, and drop any that does not

In my version

• Claude Code on a daily 8am schedule, a message only when something changed

Ask first

• Senders or kinds of mail never to draft for? (default newsletters and automated mail)
• Writing habits, and do you have a year of sent mail for the voice? (default learned from sent mail; if under a year, write ten sample replies)
• How many drafts per run, and how often? (default 6, about 9 times a day)
• May it read your calendar to propose times? (default yes, free and busy only)
• Where do questions go? (default a card in your assistant; a message if none)

Needs

• A mailbox where it can create and delete drafts but never send. Connecting: "It writes drafts and removes them once you reply; it cannot send."
• Attachment text when a reply depends on it.
• Calendar free and busy.
• A store for one verdict per message and the voice profile.
• If your assistant can only chat: paste the thread and say "draft my reply"; steps 4 to 6 apply and you paste the draft into mail yourself.

Build

1. Read: about nine times a day, inbound mail since the last run. Per message, in order: blocklist; mass mail (list headers, tracking links from a sender never replied to); calendar invites; thread already answered; from self; senders who never get replies; then one model judgement. Stop at the draft cap.
2. Decide (sender history): 10 or more messages with no reply ever means broadcast, skip; two or more replies means trusted, and the tracking-link rule is waived.
3. Memory: one stored verdict per message. A model failure retries next run rather than counting as no reply.
4. Decide (reply): from the email, the thread, earlier mail to that sender and attachment text, the model returns a draft or a question. Personal choices, commitments and decisions not in the thread become a question. Draft: "yes, Tuesday works" when Tuesday is free. Question: "will you speak at our event?"
5. Decide (second pass): check the draft against the same material; a draft with an unsupported claim is held and not saved.
6. Style: advance the thread, never bounce it back; mirror the incoming length; match the register; no manufactured closers or signature. Scheduling: two real free windows plus one fallback weekday.
7. Deliver: a draft on the thread, or a question card. Fallback: the draft as a message. Delete the draft once the person has replied in that thread.
8. Weekly, retrain the voice from 24 months of sent mail, sampled evenly per month. About 270 runs a month.

Check

• Send yourself a test thread with a factual question and run: a draft appears, nothing is sent.
• A test asking for a personal choice: a question card, no draft.
• Quiet run: a run with only newsletters creates nothing and logs its verdicts; the first real run shows in that log. Reply by hand and the draft is gone next run.

In my version

• Gmail drafts by Q, my assistant, 9 times a day, up to 6 a run; question cards on Q's home screen; voice retrained weekly from 24 months of sent mail; Google Calendar for times.

Ask first

• Where is your deals list? (a customer tool, a spreadsheet)
• What stages count as committed? (default the last two)
• When is the forecast call?

Needs

• Read access to deals, mail and calendar, or a pasted export
• If your assistant cannot read one of these, paste it when it asks. A thinner version still works
• Any password or token is entered by you into your device's secure store, never pasted into a chat or saved in a plain file

Build

1. The morning of the call, list every open deal with stage, value, close date and last contact.
2. Mark anything with no contact in nine days or a moved close date.
3. Add up committed and best case.
4. Pick three to discuss with one line each on why.

Check

• Run it on last week's list; totals match your tool
• Confirm a stuck deal is marked with its days of silence
• Say whether the writing step ran for real or on sample text. Check that every figure, quote and link matches its source, and drop any that does not

In my version

• Claude Code on a schedule, a page in my notes before each call

Ask first

• Where can it see usage or activity?
• How soon before renewal should it look? (default 90 days)

Needs

• Read access to usage and mail, or a weekly export you paste in
• If your assistant cannot read one of these, paste it when it asks. A thinner version still works
• Any password or token is entered by you into your device's secure store, never pasted into a chat or saved in a plain file

Build

1. Each morning, compare each account's recent activity to its last 60 days.
2. Flag a rise or fall of 25% or more against the 60-day average, with the number and the date.
3. Match to the renewal date.
4. Send nothing on a day with no change. Save each flagged account with the date, and do not flag it for the same reason again within 14 days.

Check

• Run it on an account that did expand; it is caught with the number
• Confirm a flag appears once, not every day
• Say whether the writing step ran for real or on sample text. Check that every figure, quote and link matches its source, and drop any that does not

In my version

• Claude Code on a daily 8am schedule, a message only when something changed

Ask first

• Which calls count? (default any calendar event with 'renewal' or the account name)
• Which mailboxes and notes hold the history?
• How far back? (default twelve months)

Needs

• Read access to email, notes and the calendar, or paste the thread history
• If your assistant cannot read one of these, paste it when it asks. A thinner version still works
• Any password or token is entered by you into your device's secure store, never pasted into a chat or saved in a plain file

Build

1. The morning of the call, collect every email and note with that account's people for twelve months.
2. Pull out each sentence where someone on our side commits to do something, with the date and who said it.
3. Check each against later messages. Mark kept, open or broken, with the evidence.
4. Show open and broken first. Never contact anyone.

Check

• Run it on a renewal you remember; every promise you recall is on it
• Confirm each open item links to the message where it was promised
• Say whether the writing step ran for real or on sample text. Check that every figure, quote and link matches its source, and drop any that does not

In my version

• Claude Code on a schedule, a page in my notes before each renewal call

Ask first

• Which accounts should it watch for, and what does each care about? (default your ten biggest, with one line each)
• Which competitors and news sources? (default five competitors and the trade press for your accounts' industry)
• What counts as relevant? (default a change an account would ask you about within a month)

Needs

• Read access to the account list, or a short list you paste in
• An assistant that can read news and competitor pages each morning and keep what it has already shown you
• If your assistant cannot read one of these, paste it when it asks. A thinner version still works
• Any password or token is entered by you into your device's secure store, never pasted into a chat or saved in a plain file

Build

1. Each weekday morning, read new headlines and the competitors' pages and compare to the saved copy.
2. Match each item to the accounts it could matter to, from what each account cares about.
3. For each match, write what it means for a conversation and one line you could say. Link the source.
4. Skip anything already shown. If nothing matches, send nothing.

Check

• Run it on last week's news; each item names an account it matters to and links its source
• Confirm an item that matches no account is dropped
• Confirm a quiet day sends nothing
• Say whether the writing step ran for real or on sample text. Check that every figure, quote and link matches its source, and drop any that does not

In my version

• Claude Code on a weekday 7am schedule, a message before my first call

Ask first

• Which metrics do you own? (default the five you look at most)
• What counts as unusual? (default beyond 15% of the four-week average)
• Where do the numbers live and what changed log should it read? (a sheet, an analytics tool; a launch or campaign calendar)

Needs

• Any password or token is entered by you into your device's secure store, never pasted into a chat or saved in a plain file
• Read access to the numbers, and a record of what shipped or ran each day
• If your assistant cannot read one of these, paste it into the chat when it asks. A thinner version still works

Build

1. Each Monday, pull the last week and the four-week average for each metric.
2. Flag any beyond the limit. For each, look at what else changed that week and name the most likely cause, marking it a guess.
3. For each likely cause you guessed, write one line saying what would show the guess was wrong, and by when, so the guess can be checked later. Do this only for a metric beyond the limit, never on a quiet week.
4. End with one thing to check. If nothing moved, say so in one line.

Check

• Run it on a past week where you know why a number moved; the likely cause matches
• Confirm a guess is labelled as a guess
• Say whether the writing step ran for real or on sample text. Check that every quote and link on the page matches its source, and drop any that does not

In my version

• Claude Code on a Monday schedule, a note in my docs and a message

Ask first

• Which accounts? (default your five biggest)
• Where do notes live?

Needs

• Read access to notes, email and deals, or paste them in
• If your assistant cannot read one of these, paste it when it asks. A thinner version still works
• Any password or token is entered by you into your device's secure store, never pasted into a chat or saved in a plain file

Build

1. List the contacts and their roles from the last 90 days of mail and notes.
2. Write what they care about in their own words.
3. Summarise what we have done and what is open.
4. Propose one move with a date.

Check

• Run it on an account you know; nothing is wrong or invented
• Confirm a person who left is not listed as current
• Say whether the writing step ran for real or on sample text. Check that every figure, quote and link matches its source, and drop any that does not

In my version

• Claude Code on a quarterly schedule, a page per account

Ask first

• Which meetings count? (default any event with an attendee whose email domain is not yours)
• How long before should it arrive? (default 30 minutes)
• Where should it land? (default a page in your notes, linked from the event)
• Which sources may it read about the attendees? (default your own mail and notes, plus their public profiles)

Needs

• Read access to your calendar, inbox and notes
• A way to write a page and attach its link to the event
• If your assistant cannot read one of these, paste it when it asks. A thinner version still works
• Any password or token is entered by you into your device's secure store, never pasted into a chat or saved in a plain file

Build

1. Every 15 minutes, look at the next 45 minutes of your calendar for events with an outside attendee that have no brief yet. Outside means an email domain that is not yours.
2. For each, find the last three threads and notes involving those people.
3. Write a page: who is in the room, where you left it, what each side owes (read from your promises list if you keep one), anything new about them, one open question.
4. Keep it under 150 words. If you have no history, say so and give one fact about the company.
5. Never contact the attendees.

Check

• Run it on a past meeting you remember; it names the thing you promised and did not do
• Confirm no sentence states something it could not find in your mail or notes
• Confirm an internal-only meeting gets no brief
• Say whether the writing step ran for real or on sample text. Check that every figure, quote and link matches its source, and drop any that does not

In my version

• Claude Code on a 15-minute schedule, a page written into my notes thirty minutes before each external meeting

Ask first

• Which calls count? (default any call you have notes for)
• Whose voice should the draft use? (default yours, learned from your last 20 sent emails)
• How soon after the call? (default within two hours, checked hourly on weekdays)

Needs

• Your notes from each call, or a written-up recording, and a way to find the email thread for the person
• Permission to create a draft only, never to send
• If your assistant cannot read one of these, paste it when it asks. A thinner version still works
• Any password or token is entered by you into your device's secure store, never pasted into a chat or saved in a plain file

Build

1. Every hour on weekdays, find calendar events that ended in the last two hours and have notes. A call has ended when its calendar event has ended.
2. Find the email thread with the attendees. If there is none, start a new draft addressed to them.
3. From the notes, list what was agreed, who owns each item and by when, and one open question. If the notes say nothing was agreed, say so on the first line of the draft and do not invent next steps.
4. Write a short reply in your voice as a draft on that thread. Never send it.
5. Add each commitment you made in the draft to the promises list, so it is tracked from the moment you hang up.

Check

• Run it on yesterday's call; every commitment in the draft appears in your notes
• Confirm the draft is in drafts and nothing was sent
• Confirm each commitment also appears on the promises list
• Say whether the writing step ran for real or on sample text. Check that every figure, quote and link matches its source, and drop any that does not

In my version

• Claude Code on an hourly weekday schedule, a draft left on the existing thread and each commitment added to my promises list