Turn a manual sales process into an automation blueprint
Converts a fuzzy 'we do this by hand every week' description into a structured blueprint with a detectable trigger, labeled steps, edge cases, human checkpoints, and failure alerts. It forces the design decisions people usually discover mid-build, so what you hand to n8n, Zapier, or an ops hire is assembly work, not archaeology.
You are an automation architect who has systemized outbound operations for over 100 B2B teams. Your rule: document the process before you touch a tool, because automating a broken process just makes bad things happen faster. I'm going to describe a manual process my team runs today. Turn it into a build-ready automation blueprint with this exact structure: 1. TRIGGER — the event that starts the workflow, phrased as something a system can detect (not 'when someone remembers'). 2. STEPS — a numbered sequence, each step labeled [AUTOMATE], [AI], or [HUMAN], with the data in and data out of each. 3. EXCEPTIONS — the 3-5 edge cases most likely to break the flow and how each is handled. 4. HUMAN CHECKPOINTS — where a person must review before the flow continues, and exactly what they check. 5. TOOLS — a suggested stack using tools I already have first; recommend new tools only with a one-line justification. 6. FAILURE ALERTS — what gets flagged, and where, when something silently stops. Quality bar: no step may say 'the system handles it' without naming the mechanism. Never assign [AI] to a judgment call that needs a human. If a step in my process shouldn't exist at all, say so and delete it. Before you design anything, interview me. Ask these questions ONE AT A TIME, waiting for my answer each time: 1. Walk me through the process step by step as it happens today — who does what, in which tool? 2. How often does it run, and roughly how long does each run take? 3. What tools are in your stack right now (CRM, sequencer, enrichment, automation platform)? 4. Where does it break most often, and what happens when it does? 5. What must a human always approve, no matter what? Once you have my answers, produce the blueprint. If any step is vague, ask one follow-up before locking it in.
How to use it
- 1
Copy the prompt into Claude, ChatGPT, or any LLM.
- 2
Answer the five interview questions honestly — especially question 4, because the breakage points shape the exception handling.
- 3
Review the blueprint and challenge any step labeled [AI] that involves judgment you wouldn't delegate to a new hire.
- 4
Hand the finished blueprint to whoever builds it, or feed it into the n8n workflow design prompt from this category.
Best practices
Describe the process as it actually happens, including the workarounds — the ugly parts are where automations die.
If the blueprint has zero [HUMAN] steps, push back; fully autonomous flows around revenue almost always need at least one checkpoint.
Run the prompt once per process, not for your whole operation at once — blueprints degrade when they try to cover everything.
Save the final blueprint in your ops docs; it doubles as the maintenance runbook when the automation breaks.
Example: what this looks like in practice
A RevOps lead at a 60-person staffing firm describes their weekly list-building ritual: an SDR exports search results, pastes them into a sheet, checks each company's careers page, and uploads survivors to their sequencer. The interview surfaces that the careers-page check is the judgment step and that broken runs go unnoticed for days. The blueprint comes back with a scheduled trigger, a scraper step labeled [AUTOMATE], an AI relevance check labeled [AI] with a confidence threshold, a [HUMAN] spot-check of 10 rows per batch, and a Slack alert when row count drops below 50. Their VA builds it in n8n in an afternoon, reclaiming roughly six SDR hours a week.
Best fit
This prompt is one gear in a bigger machine. We orchestrate 20+ tools into outbound systems our clients own — and guarantee the results.
Apply for a Pilot Spot → →Frequently asked questions
Start with high-frequency, low-judgment tasks: list building, enrichment, CRM updates, routing. This prompt makes the decision mechanical — any step labeled [AUTOMATE] with high weekly volume is a candidate, while steps that stay [HUMAN] after honest review are not. Ranking blueprints by hours saved per week gives you a build order in minutes.
More automation & workflow design prompts
Build a system prompt for an AI cold email reply agent
Interviews you about your offer, booking flow, and voice, then generates a complete production system prompt for an AI reply agent — classification taxonomy, response rules, escalation conditions, and a strict JSON output contract an n8n or Zapier workflow can parse. You get the artifact teams pay thousands to have written, tuned to how you actually talk.
Architect a Clay table before you burn the credits
Designs the Clay table before you build it: sources, column order, run conditions, waterfalls, outputs, and a credit estimate. Column ordering and gating are where most teams silently overspend, so the prompt enforces filter-first architecture and makes every paid column defend its run condition — typically cutting credit burn 40-70% versus enrich-everything builds.
Audit your full sales tool stack for overlap, gaps, and waste
Turns your tool list into a capability matrix that exposes overlaps, gaps, and zombie subscriptions, then delivers cut/keep/consolidate verdicts with savings math and safe migration notes. Teams typically find 20-35% of GTM tool spend is redundant — this makes the case with a matrix, not a hunch.