playbooks

FIRE CEO Playbook: Run a Company From a Terminal

The CEO operating system inside Claude Code. Pipeline, content, outreach, hiring, billing: every CEO task as a slash command.

18 min read·

Playbook · 6 CEO Domains · FIRE Framework

FIRE CEO Playbook: Run a Company From a Terminal

The average CEO spends 23 hours a week on coordination overhead: status checks, follow-ups, content approvals, invoice chasing, hiring updates, team unblocking. None of that is the job. The job is decisions and direction. Everything else should run from a command.

This playbook maps every repeating CEO motion to a slash command. Pipeline, content, outreach, hiring, billing, people: six domains, six command sets, one terminal. Based on the FIRE framework and patterns from 23 company engagements.

The goal is not to automate your judgment. It is to remove the coordination tax so the only thing left on your calendar is judgment.

mindset

Domains 1-3

execution

Domains 4-5

scale

Domain 6

Anthropicn8nSlackLinearSupabaseHubSpotClayApollo
01
mindset

Pipeline

Your pipeline is a process, not a spreadsheet. If your CRM requires manual entry, you don't have a pipeline; you have a graveyard of good intentions.

Most CEOs manage pipeline through a combination of memory, Slack threads, and weekly calls with the sales team. That loop has a latency of days. By the time you know a deal is cold, the opportunity has already moved. The terminal CEO sees pipeline state in real time (deal stage, last contact, next action, and risk score) without opening a CRM.

Every pipeline motion becomes a slash command. You don't open HubSpot to check status. You run /pipeline and get a ranked list of deals with next actions. You run /deal-risk and surface the three deals most likely to slip this week. The system does the CRM work. You do the thinking.

Sales reps spend 65% of their time on non-selling activities. Most of it is updating CRM records.

DO THIS

Map every pipeline touchpoint that requires manual data entry. Replace each one with a command that pulls data from where it already lives: email, calendar, call transcripts.

Slash Commands

/pipelinePull live deal list ranked by close probability
/deal-riskSurface deals at risk of slipping this week
/follow-up [name]Generate personalized follow-up from last call transcript
/pipeline-briefWeekly pipeline summary for team standup
Skills:clay-outreachpersonalized-value-outreachMCPs:HubSpot MCPSlack MCP
02
mindset

Content

Content is not a marketing function. It is a trust machine. Every post, article, and playbook you don't publish is a conversation you didn't have.

The CEO who publishes consistently wins the trust game before the sales conversation starts. But most CEOs treat content as something that happens when there's time, which means it never happens. The terminal CEO has a content system that runs from a single command, turning raw thinking into published assets.

One LinkedIn post becomes a carousel, a newsletter section, and three outreach hooks. One client story becomes a case study, a playbook chapter, and a webinar talking point. The content loop is not about creating more. It's about extracting more value from what already happened. Your calendar is full of content. The command extracts it.

CEOs who publish weekly on LinkedIn generate 3x more inbound than those who post monthly. The gap is consistency, not quality.

DO THIS

Run /content-brief every Monday with your top 3 insights from the previous week. Let the system turn those three sentences into a week of content assets.

Slash Commands

/content-briefTurn raw insights into a week of content assets
/post [topic]Draft a LinkedIn post with hook and structure
/repurpose [url]Extract carousel, newsletter, and hooks from existing content
/newsletter-draftGenerate newsletter from weekly call notes
Skills:linkedin-hookslinkedin-automationgeneral-blog-postMCPs:Notion MCPSlack MCP
03
mindset

Outreach

Personalized outreach at scale is not spam. It is the ability to say the right thing to the right person at the right time, without hiring a team of SDRs.

Cold outreach fails because it is generic. The message that goes to 500 people reads like it was written for no one. The terminal CEO runs outreach that reads like it was written for one person, because it was, by a system that researched that person in 90 seconds before drafting the message.

The /outreach command pulls the prospect's recent posts, their company's latest news, and their likely pain point based on ICP scoring. It drafts a message that references something specific. The conversion rate on specific messages is 4-8x higher than generic ones. You can send 10 targeted messages in the time it takes to send 50 generic ones, and get better results.

Personalized outreach has a 29% higher open rate and 41% higher reply rate than templated sequences.

DO THIS

Replace your current outreach sequence with /outreach [name]. Benchmark reply rate over 30 days. If it doesn't double, the research layer needs work.

Slash Commands

/outreach [name]Research prospect and draft hyper-personalized first message
/icp-score [company]Score a company against your ICP criteria
/sequence [name]Build a 3-touch sequence with context from previous messages
/reply-draftDraft reply to inbound message with context from CRM
Skills:clay-outreachpersonalized-value-outreachMCPs:Gmail MCPSlack MCP
04
execution

Hiring

The job posting is not the bottleneck. The evaluation loop is. Most companies spend 8 weeks deciding what 8 hours should have answered.

A company walked in with a 3-month hiring process for an engineer. The first interview was a culture screen. The second was a resume review. The third was a technical challenge. The fourth was a team fit. At no point did anyone test whether the person could do the job. They were optimizing for signals that had nothing to do with the work.

The terminal CEO runs hiring as a command loop. /job-post generates the role brief from a three-line description. /screen [applicant] scores the resume against criteria. /challenge [role] generates a work-sample test specific to the real work. The evaluation loop collapses from weeks to days without losing signal quality, because the signal was always in the work, not the interview.

Work-sample tests predict job performance 5x better than unstructured interviews. Most companies still use unstructured interviews.

DO THIS

For your next hire, replace all interviews before the work-sample test with /screen. Run the challenge first. Interview only the people who pass it.

Slash Commands

/job-post [role]Generate role brief and requirements from three-line description
/screen [applicant]Score resume against role criteria and flag gaps
/challenge [role]Generate work-sample test aligned to actual job tasks
/hiring-pipelineTrack all active candidates with stage and next action
Skills:scaffoldplanMCPs:Notion MCPGmail MCP
05
execution

Billing

Revenue leaks silently. The deals that closed, the hours that shipped, the invoices that were never sent. By the time you notice, the quarter is already wrong.

One company had 14% of their billed work going uninvoiced because the billing system required engineers to log hours manually and sales to approve the invoice before it went out. Engineers don't log. Sales misses approvals. Finance chases both. The revenue was there. The loop to capture it was broken.

The terminal CEO runs billing from a command. /billing-status shows every open engagement, what has shipped, what is invoiced, and what is overdue. /invoice [client] generates the invoice from project records and sends it. The system knows what shipped because it's connected to where the work lives, not a separate manual entry point.

Small businesses lose an average of 13% of annual revenue to invoicing delays and errors. Most of it is recoverable with a closed billing loop.

DO THIS

Audit last quarter. How many engagements had a gap between delivery and invoice date? Anything over 7 days is a broken loop. Map the handoffs and replace each one with a command.

Slash Commands

/billing-statusShow all open engagements with invoice and payment status
/invoice [client]Generate and send invoice from project delivery records
/overdueSurface overdue invoices with recommended follow-up action
/revenue-briefMonthly revenue summary with collected, outstanding, and at-risk
Skills:auditqaMCPs:Gmail MCPGoogle Calendar MCP
06
scale

People

Your best people are building on the side. They stopped asking permission because your decision loop is too slow. The terminal CEO runs people ops from a command, and the loop closes before they leave.

They are automating reports on personal laptops. Testing models with their own accounts. Building shortcuts that make their work 3x faster, shortcuts they cannot share because there is no sanctioned way to do it. The gap between what your best people know is possible and what the company is doing grows every month. That gap becomes frustration. Frustration becomes quiet quitting. Quiet quitting becomes a resignation email.

The terminal CEO runs /team-pulse to surface blockers before they become reasons to leave. Runs /sandbox-request to spin up a testing environment before anyone has to ask twice. Runs /weekly-brief to generate a one-paragraph summary from each person's async updates. The people loop closes in hours, not weeks. And the people who close loops stay.

Employees who feel their ideas are acted on within a week are 3x more likely to stay than those who wait a month.

DO THIS

Run /team-pulse every Friday. Ask three questions: what shipped, what's blocked, what should we try next week. If the same blocker appears twice, it is a system problem, not a people problem.

Slash Commands

/team-pulseWeekly async check-in with blockers, wins, and next priorities
/sandbox-requestSpin up testing environment and assign access in one command
/weekly-briefGenerate team summary from async updates for leadership
/retro [sprint]Run structured retrospective and extract learnable patterns
Skills:task-statuspipeline-viewMCPs:Slack MCPNotion MCP

Daily & Weekly Cadence

The CEO terminal runs on a cadence, not a calendar. These are the commands that fire on schedule.

Daily

/pipelineStart of day: know your deals
/team-pulseCheck blockers before they compound
/post [insight]One idea → one LinkedIn post
/overdueEnd of day: close any open billing

Weekly

/pipeline-briefMonday: full pipeline review with team
/content-briefMonday: extract week's content from last week's calls
/revenue-briefFriday: collected, outstanding, at-risk
/retro [week]Friday: patterns, blockers, next priorities
/hiring-pipelineFriday: candidate status and next interviews
/weekly-briefFriday: async team summary for leadership

6 Domains at a Glance

mindset

  • 01Pipeline
  • 02Content
  • 03Outreach

execution

  • 04Hiring
  • 05Billing

scale

  • 06People

Ready to run your company from a terminal?

DM Walid on LinkedIn or book a 1-hour analysis. No decks. No fluff. Just the loop.

Want this running in your stack?

AY Automate builds AI automation systems for production teams.