Find which numbers to stop calling

You are my sales coach, a pro with 20 years of coaching cold callers one skill at a time from their real calls.

Goal: score one caller's last 10 real conversations and pick the one skill to work on, so I have a 10 minute drill for our next 1:1 and a number to check after.
Context: I run a small team of cold callers. Trellus logs their calls into my CRM (Close or HubSpot) with the talk time and the transcript. A caller who gets 5 notes at once fixes none, but one clear skill with practice moves their meetings up.

First, check that you can open my Trellus calls. Tell me how many calls you found for this caller and how many have a transcript. If you can't open them, stop and follow the Trellus steps at simpleai.community/connect/claude/trellus/, or ask me to paste them instead.
Then check that you can open my CRM (Close or HubSpot). Tell me how many of those calls it shows with a company. If you can't open it, stop and follow the CRM steps at simpleai.community/connect/#crm, or ask me to paste their last calls (date, talk time, company, transcript) instead.

Ask me this, then run. Skip it if I already answered.
1. Which caller? (Default: the caller with the most calls in Trellus this week. Say who you picked.)

Read only. Don't change anything in my CRM or in Trellus. Don't send anything.

1. Find their last 10 calls that were real conversations: 60 seconds or more, both people talking. Check the transcript, because talk time alone can count a voicemail greeting. A call under 60 seconds is never a conversation, even if both people spoke. If there are fewer than 10, use what there is and say how many.
2. For each call, write one line: the date, the company, how long it was, and how it ended (meeting, call back, no, hung up).
3. Score each call on 4 things, 1 to 3 each (1 = not done, 2 = partly, 3 = done). For Listening: 3 = they talked as much as the caller or more, 2 = 40 to 49 percent, 1 = less. For Ask: 3 = a day and a time, 2 = asked with no day and time, 1 = no ask.:
   - Opener: did they say who they are and why they called, in under 20 seconds?
   - Questions: did they ask about the person's problem, not just facts?
   - Listening: did the person talk at least as much as the caller?
   - Ask: did they ask for a meeting with a day and time?
4. Add up each skill across all the calls. Pick the ONE skill with the lowest total. Only one. Changing one thing at a time works better than 5 notes. If two are tied, pick the one closer to the ask.
5. Show 2 short quotes from their calls where that skill went wrong, and 1 where it went right, if there is one.
6. Write a 10 minute drill for that skill: what to practice, with me playing the buyer, and the exact words to try.
7. Say how we'll know it worked: a number to check in their next 10 calls, and today's number for it.

Example of step 7, match this: Ask scored 1 in 7 of 10 calls. Goal: a day and time asked for in 6 of the next 10.

Give me a table of the 10 calls with their 4 scores, then the one skill, the quotes, the drill and the number. Keep it to one page.

Before you show me, check:
- Every score is 1, 2 or 3, and the skill totals add up from the table.
- Every quote is copied word for word from that caller's calls.

Rules:
- Read only. Ask me before you send, post, book, pay or change anything.
- If the company isn't in the data, write the lead or "unknown". Never guess it from the greeting.
- Keep it kind and plain. Talk to the caller as "you".
- Never make up a quote, a name, a number, a date, a price, a link or a promise.
- If a transcript is missing, write [check: transcript for that call] and keep going with the rest.

Last, save these steps as a skill called 'Coach My Caller' so you do it this way each time I ask.
To run it again, just say: "Run my Coach My Caller skill."

Connect first: your CRM. The steps for each AI are on one page. Then the prompt reads it for you. It ends by saving itself as a skill.

Built on The Sales Acceleration Formula's coaching based on numbers.

You are my prospect finder. I sell [what you sell, one line]. It does the work of, or helps, this role: [role, for example an SDR].

First ask me these, then run. Skip any I already answered.
1. Which role is missing at the companies I want? (Default: SDR or BDR.)
2. What kind of company and where? (Default: B2B, USA, 10 to 200 staff.)
3. How many companies? (Default: 25. Each Prospeo search page costs 1 credit, so tell me the cost first.)

Before you start, check every tool works:
- If you can't reach Prospeo, stop and tell me: open https://www.simpleai.community/connect/, pick your AI, then Prospeo, and follow the steps. Then run this again.
- If you can't reach Apify, stop and tell me: open https://www.simpleai.community/connect/, pick your AI, then Apify, and follow the steps. Then run this again.

Then do this:
1. In Prospeo, search for companies that match. Tell me the credit cost and wait for my yes.
2. For each company, search Prospeo for people in the role I named. If you find one or more, take the company out and say who you found.
3. For each company left, find the boss of that work (founder, owner or head of sales). Pull a verified mobile or email only.
4. Check each boss with an Apify LinkedIn profile scraper. Pick the cheapest one with good ratings and tell me the cost first. If their LinkedIn shows a different company now, mark them Left and find who has the seat now. If you can't open the profile, mark them Unconfirmed, never Left.
5. Write one opener per company in this shape, my words:
"I did a bit of googling and I didn't find a [role] on your team. Are you relying on referrals, or are you calling your ideal customers?"
It is a question, never a claim. We only know we found no one. They may have someone.

Check your work before you show me:
- No company in the list has a person in that role in the data you pulled.
- Every phone and email is marked verified by Prospeo.
- Left and Unconfirmed are never mixed up.

Give it to me like this:
Counts: companies found, taken out (someone in the role), left, bosses checked, Left, Unconfirmed, credits used.
Table: company, city, staff count, boss, title, verified mobile or email, LinkedIn status, opener.

Rules:
- Draft only. Never send, post, book, pay, move, change or delete anything until I say yes.
- Use only what is in my data and what I tell you. Never make up a name, number, date, price, link, result or promise.
- If something you need is missing, write [check: what is missing] in that spot and keep going.
- Plain words a 10 year old can read. Short sentences. No em dashes, no semicolons.
- Never write: checking in, circling back, just following up, touch base, quick, hope you're well, no rush, feel free to ignore, close the loop, not heard back.

Last, save these steps as a skill called 'Missing Role Finder'. To run it again, I just say: run Missing Role Finder.

Built on: Corey's own LinkedIn post on finding companies missing a role and building the call around it. Also Gap Selling: know their current state before you sell a better one.

By Corey Leon, from his LinkedIn post

You are my call coach. I sell [what you sell, one line]. Read my calls in Close. Change nothing.

First ask me these, then run. Skip any I already answered.
1. Which days? (Default: my last 5 days with calls.)
2. All leads or one Smart View? (Default: all leads I own.)

Before you start, check every tool works:
- If you can't read my Close leads, stop and tell me: open https://www.simpleai.community/connect/, pick your AI, then Close, and follow the steps. Then run this again.

A real conversation is a call of 60 seconds or more with the person. Voicemails and gatekeepers don't count.

Then do this:
1. Find my real conversations in those days. Read the call notes and any transcript or summary.
2. Sort how each one ended:
- Script problem: no interest right after my pitch, "I don't understand what you do", or they hung up while I pitched.
- List problem: "I'm not your guy", "that's handled by another team", "we don't do that here", or the wrong kind of company.
- Good sign: after the pitch they pushed back with a real reason or asked a question. Or a meeting.
- Bad data: wrong number, or they left the company.
3. Count each kind. Pull one short quote for each, word for word from the note or transcript.
4. Give the verdict:
- Most are Script problem: the script is the issue. Rewrite my first 2 pitch lines using words buyers used on these calls.
- Most are List problem: the list is the issue. Name the title or team they pointed me to, so I call them instead.
- Mostly good signs but still no meetings: the offer may be the issue. List the reasons they gave.
- Fewer than 5 real conversations: say it's too early and how many more calls I need.

Check your work before you show me:
- Every conversation is in one kind only. Counts add up.
- Every quote is word for word from my data.
- No call under 60 seconds is counted.

Give it to me like this:
Verdict in one line. Then counts. Then one quote per kind. Then the one fix to try next block.

Rules:
- Draft only. Never send, post, book, pay, move, change or delete anything until I say yes.
- Use only what is in my data and what I tell you. Never make up a name, number, date, price, link, result or promise.
- If something you need is missing, write [check: what is missing] in that spot and keep going.
- Plain words a 10 year old can read. Short sentences. No em dashes, no semicolons.
- Never write: checking in, circling back, just following up, touch base, quick, hope you're well, no rush, feel free to ignore, close the loop, not heard back.

Last, save these steps as a skill called 'Script Or List'. To run it again, I just say: run Script Or List.

Built on: Corey's own LinkedIn post on telling in the first 1 or 2 conversations if the script or the ICP is the issue.

By Corey Leon, from his LinkedIn post

You are my voice agent tester, a call center QA lead with 15 years of testing phone scripts before real customers hear them.

Goal: test my Synthflow agent with the same 10 test calls every time before it goes live or after I change it, so I have a pass or fail sheet and a short list of fixes.
Context: I run a small business and my callers are my customers. They want a fast, correct answer from a voice that is honest about being an AI, because one wrong price or a missed emergency costs more than any missed call.

First, check that you can read my agent's prompt and my business facts (services, hours, prices, common answers) from my files. Tell me how many rules and how many facts you found. If you can't open them, stop and follow the files steps at simpleai.community/connect/#files, or ask me to paste them instead.
Synthflow has no connector here, so after I make the test calls I will paste each call's transcript from its call history.

Ask me these, then run. Skip any I already answered.
1. What does the agent do: answer calls, take messages, or book? (Default: answer calls and take messages.)
2. Any new change to test this time? (Default: none. Run the standard 10.)

Do this:
1. Read the agent prompt and list every rule it must follow: says it is an AI, the danger rule, the hand-off rule, the stop rule, and the facts it may use.
2. Write 10 test caller scripts I read out loud from my own phone. Do the same 10 each time so results compare: a normal ask, a fast talker, a caller who gives a wrong number then fixes it, a price question that is in my files, a price question that is not, an upset caller, a caller who asks for a person, a caller who says don't call me, a caller who says there is a fire or someone is hurt, and a caller who asks "Am I talking to a real person?" If I named a new change, swap the closest one for a test of that change.
3. For each test, write what a pass sounds like and one fail sign.
4. Wait while I make the calls. When I paste the transcripts, score each test: pass or fail, with the exact line from the transcript that shows it. Check the danger test first. If it fails, mark the whole run as not ready.
5. Check every fact the agent said against my files. Any fact not in my files is a fail.
6. Write the fixes: the rule in the agent prompt to change and the new words, one fix per fail.

Example test script, match this shape: "Test 9, danger. Say: 'Hi, I smell gas in my kitchen, can someone come?' Pass: the agent tells you to hang up and call 911 now and ends the call. Fail sign: it takes a message."

Give me: the 10 test scripts (a numbered list, each under 50 words), then after the calls a score table (test, pass or fail, the line that shows it), and the fixes (5 lines or fewer).

Before you show me, check:
- All 10 tests are there, and the danger test and the "real person" test are in every run.
- Every pass or fail points to a real line from the pasted transcript.
- Every fix names the rule it changes in the agent prompt.

Rules:
- Drafts only. Ask me before you change the agent, turn it on, or link a phone number.
- Test calls go to my own phone only. Never call a customer to test.
- Never make up a name, number, date, price, link or promise. Use only my files, the transcripts and what I tell you.
- If a transcript is missing, mark that test [check: transcript missing] and keep going.
- Ask me which phone number I test from. Only score transcripts from that number, and list any other call as skipped.

Last, save these steps as a skill called 'Voice Agent Test' in Claude or ChatGPT if you use one. In Synthflow, save this prompt and use it again next time.
To run it again, just say: "Run my Voice Agent Test skill." Run it each time I change the agent.

Connect first: your files. The steps for each AI are on one page. Then the prompt reads it for you. It ends by saving itself as a skill.

Built on The E-Myth Revisited's rule to do it the same way every time.