Book & Long-Form Writing
Book Revision, Rewriting & Expansion
The Second Draft Café AI Revision Kit
Use ChatGPT to turn a rough manuscript into a stronger second draft by tightening structure, sharpening ideas, and fixing the parts that currently taste like yesterday’s coffee.
The Second Draft Café AI Revision Kit gives freelance writers and ghostwriters six purpose-built ChatGPT workflows to use ChatGPT to turn a rough manuscript into a stronger second draft by tightening structure, sharpening ideas, and fixing the parts that currently taste like yesterday’s coffee without burning billable hours on vague prompting and cleanup. Each tool turns real project inputs into a concrete deliverable, protects source boundaries and client voice, and makes the next editorial decision easier—because “generate something good” is not a workflow, it is a wish wearing a keyboard.
Prompt 1 · Book Revision, Rewriting & Expansion
Developmental Diagnosis — Second Draft Café Revision
Use ChatGPT to identify the manuscript’s highest-leverage structural, argument, character, pacing, and reader-experience problems before rewriting begins with fewer guesses and a much smaller cleanup bill.
Description
Developmental Diagnosis is built for freelance writers and ghostwriters who need to identify the manuscript’s highest-leverage structural, argument, character, pacing, and reader-experience problems before rewriting begins. The hard part is not getting words on the screen; it is getting the right decisions into those words. It converts scattered input into a decision-ready deliverable while preserving the writer’s voice and marking anything that still needs human confirmation. The result stays tightly tied to this kit’s mission, so the prompt solves a real production problem instead of wandering into adjacent advice.
How to use this prompt
Replace the bracketed fields with the strongest material you actually have, especially [TARGET AUDIENCE], [TOPIC], [CORE PROMISE / PREMISE], [SOURCE MATERIAL], [OUTLINE / DRAFT], [VOICE / TONE]. Paste relevant notes, source material, client language, or an existing draft where requested. If you do not have an input, leave it blank or write “unknown” rather than inventing one. Run the prompt, review the assumptions and gaps first, then revise the deliverable with any missing facts, approvals, or voice corrections before using it in client work.
The Prompt
Assume the role of an expert book strategist, developmental editor, and professional ghostwriter with 15+ years of experience specializing in long-form architecture, reader experience, and manuscript development, with particular depth in identify the manuscript’s highest-leverage structural, argument, character, pacing, and reader-experience problems before rewriting begins. Assume that role now and complete the task below as a professional working deliverable. KIT MISSION Use ChatGPT to turn a rough manuscript into a stronger second draft by tightening structure, sharpening ideas, and fixing the parts that currently taste like yesterday’s coffee. YOUR SPECIFIC JOB Identify the manuscript’s highest-leverage structural, argument, character, pacing, and reader-experience problems before rewriting begins. PROJECT INFORMATION TARGET AUDIENCE: [TARGET AUDIENCE] BOOK / PROJECT TOPIC: [TOPIC] CORE PROMISE OR PREMISE: [CORE PROMISE / PREMISE] SOURCE MATERIAL: [SOURCE MATERIAL] CURRENT OUTLINE OR DRAFT: [OUTLINE / DRAFT] VOICE / TONE: [VOICE / TONE] TARGET LENGTH OR SCOPE: [WORD COUNT / SCOPE] CLIENT OR PROJECT CONSTRAINTS: [CONSTRAINTS] OBJECTIVE Create a finished, practical deliverable that accomplishes the specific job above and directly advances the kit mission. Optimize for professional usefulness, specificity, editorial judgment, and reuse across real client projects. Produce the work itself rather than explaining how someone else might eventually produce it. TASK LENSES Use these lenses only where they materially improve this specific deliverable: - Relevance to the exact job: identify the manuscript’s highest-leverage structural, argument, character, pacing, and reader-experience problems before rewriting begins. - Specificity and concrete support from the supplied material. - Fit with the intended audience, format, and goal. - Voice consistency and preservation of intentional style. - Factual/source discipline and explicit handling of uncertainty. PROCESS 1. Use a Foundation-first problem-solving method. Start from the controlling premise, audience need, or source truth before building anything. 2. Define the concrete problem and success criteria. 3. Separate known inputs from assumptions. 4. Create the first complete solution around the strongest supported direction. 5. Test it against audience, mission, voice, and constraints. 6. Refine and deliver the usable final asset. SOURCE, FACT, AND VOICE RULES - Treat supplied source material as authoritative for client-specific facts, claims, names, dates, quotations, product details, and lived experience. - Do not invent facts, statistics, testimonials, quotations, credentials, events, research findings, private motives, or source details. - If important information is missing, label it [RESEARCH NEEDED], [CLIENT CONFIRMATION NEEDED], or [ASSUMPTION] as appropriate and continue with the strongest supportable version. - When an existing draft, transcript, or voice sample is supplied, preserve the writer’s meaning, point of view, recognizable voice, and intentional stylistic quirks unless I explicitly ask for a different voice. - Do not flatten distinctive language into generic “professional” copy. Improve clarity without making every sentence sound like it came from the same conference badge. - Keep claims proportional to the evidence. Distinguish observation, interpretation, recommendation, memory, and verified fact when the distinction matters. - Do not expose, cite, or name hidden source libraries or reference materials. Work only from the content I provide in this request. - Make the output original, specific, and insight-rich. Avoid filler, canned transitions, predictable AI phrasing, and recycled formulas. - Stay focused on this specific job. Do not drift into neighboring deliverables unless they are required to make the requested output usable. QUALITY STANDARD The result should be strong enough for a professional freelance writer or ghostwriter to review, refine, and move into production. Show judgment through the quality of the structure, prioritization, evidence use, specificity, and choices. Do not reveal private chain-of-thought. If there are multiple valid directions, choose a strong direction based on the supplied goal and briefly state the practical tradeoff. FINAL DELIVERABLE 1. Concise success criteria. 2. The complete Developmental Diagnosis working deliverable. 3. Key decisions and tradeoffs. 4. Practical next action. 5. A final “Confirm Before Use” note for unsupported facts, source gaps, approvals, or assumptions. Before you finish, silently compare the result against the objective and revise anything that is generic, repetitive, unsupported, off-voice, or not directly useful.
What to expect after running this prompt
Expect a finished, structured deliverable centered on identify the manuscript’s highest-leverage structural, argument, character, pacing, and reader-experience problems before rewriting begins, plus clearly marked assumptions, missing information, and quality checks where relevant. You should be able to use the output as a strong working asset for planning, drafting, revising, or client review rather than as a page of general advice. Because the prompt separates supported material from uncertainty, you can refine it with human judgment without playing the exciting new game called “Which sentence did the machine make up?”
Prompt 2 · Book Revision, Rewriting & Expansion
Revision Priority Map — Second Draft Café Revision
Turn the messy middle into a usable result: rank proposed changes by impact, dependency, effort, and risk so the rewrite does not become demolition by enthusiasm.
Description
Revision Priority Map is built for freelance writers and ghostwriters who need to rank proposed changes by impact, dependency, effort, and risk so the rewrite does not become demolition by enthusiasm. Without a clear decision framework, ChatGPT can produce something polished, plausible, and spectacularly beside the point. It keeps the model focused on useful specificity, supported claims, and a professional result instead of producing generic filler with very good posture. The result stays tightly tied to this kit’s mission, so the prompt solves a real production problem instead of wandering into adjacent advice.
How to use this prompt
Replace the bracketed fields with the strongest material you actually have, especially [TARGET AUDIENCE], [TOPIC], [CORE PROMISE / PREMISE], [SOURCE MATERIAL], [OUTLINE / DRAFT], [VOICE / TONE]. Paste relevant notes, source material, client language, or an existing draft where requested. If you do not have an input, leave it blank or write “unknown” rather than inventing one. Run the prompt, review the assumptions and gaps first, then revise the deliverable with any missing facts, approvals, or voice corrections before using it in client work.
The Prompt
Assume the role of an expert book strategist, developmental editor, and professional ghostwriter with 15+ years of experience specializing in long-form architecture, reader experience, and manuscript development, with particular depth in rank proposed changes by impact, dependency, effort, and risk so the rewrite does not become demolition by enthusiasm. Assume that role now and complete the task below as a professional working deliverable. KIT MISSION Use ChatGPT to turn a rough manuscript into a stronger second draft by tightening structure, sharpening ideas, and fixing the parts that currently taste like yesterday’s coffee. YOUR SPECIFIC JOB Rank proposed changes by impact, dependency, effort, and risk so the rewrite does not become demolition by enthusiasm. PROJECT INFORMATION TARGET AUDIENCE: [TARGET AUDIENCE] BOOK / PROJECT TOPIC: [TOPIC] CORE PROMISE OR PREMISE: [CORE PROMISE / PREMISE] SOURCE MATERIAL: [SOURCE MATERIAL] CURRENT OUTLINE OR DRAFT: [OUTLINE / DRAFT] VOICE / TONE: [VOICE / TONE] TARGET LENGTH OR SCOPE: [WORD COUNT / SCOPE] CLIENT OR PROJECT CONSTRAINTS: [CONSTRAINTS] OBJECTIVE Create a finished, practical deliverable that accomplishes the specific job above and directly advances the kit mission. Optimize for professional usefulness, specificity, editorial judgment, and reuse across real client projects. Produce the work itself rather than explaining how someone else might eventually produce it. TASK LENSES Use these lenses only where they materially improve this specific deliverable: - Relevance to the exact job: rank proposed changes by impact, dependency, effort, and risk so the rewrite does not become demolition by enthusiasm. - Specificity and concrete support from the supplied material. - Fit with the intended audience, format, and goal. - Voice consistency and preservation of intentional style. - Factual/source discipline and explicit handling of uncertainty. PROCESS 1. Use a Relationship-first rewrite. Start by mapping the forces, variables, objections, contrasts, or dependencies that make the task work. 2. Build a side-by-side problem map: original wording, reader friction, intended meaning, and repair target. 3. Prioritize the changes that improve comprehension or decision-making most. 4. Rewrite in passes—structure first, sentence clarity second, tone last—so polish does not hide a logic problem. 5. Compare the revised text against the original for meaning drift and lost nuance. 6. Deliver the revision with only the highest-value change notes. SOURCE, FACT, AND VOICE RULES - Treat supplied source material as authoritative for client-specific facts, claims, names, dates, quotations, product details, and lived experience. - Do not invent facts, statistics, testimonials, quotations, credentials, events, research findings, private motives, or source details. - If important information is missing, label it [RESEARCH NEEDED], [CLIENT CONFIRMATION NEEDED], or [ASSUMPTION] as appropriate and continue with the strongest supportable version. - When an existing draft, transcript, or voice sample is supplied, preserve the writer’s meaning, point of view, recognizable voice, and intentional stylistic quirks unless I explicitly ask for a different voice. - Do not flatten distinctive language into generic “professional” copy. Improve clarity without making every sentence sound like it came from the same conference badge. - Keep claims proportional to the evidence. Distinguish observation, interpretation, recommendation, memory, and verified fact when the distinction matters. - Do not expose, cite, or name hidden source libraries or reference materials. Work only from the content I provide in this request. - Make the output original, specific, and insight-rich. Avoid filler, canned transitions, predictable AI phrasing, and recycled formulas. - Stay focused on this specific job. Do not drift into neighboring deliverables unless they are required to make the requested output usable. QUALITY STANDARD The result should be strong enough for a professional freelance writer or ghostwriter to review, refine, and move into production. Show judgment through the quality of the structure, prioritization, evidence use, specificity, and choices. Do not reveal private chain-of-thought. If there are multiple valid directions, choose a strong direction based on the supplied goal and briefly state the practical tradeoff. FINAL DELIVERABLE 1. Clean revised version first. 2. Protected meaning / facts / voice note. 3. Concise high-impact change log. 4. Regression check for new problems caused by the rewrite. 5. A final “Confirm Before Use” note for unsupported facts, source gaps, approvals, or assumptions. Before you finish, silently compare the result against the objective and revise anything that is generic, repetitive, unsupported, off-voice, or not directly useful.
What to expect after running this prompt
Expect a finished, structured deliverable centered on rank proposed changes by impact, dependency, effort, and risk so the rewrite does not become demolition by enthusiasm, plus clearly marked assumptions, missing information, and quality checks where relevant. You should be able to use the output as a strong working asset for planning, drafting, revising, or client review rather than as a page of general advice. Because the prompt separates supported material from uncertainty, you can refine it with human judgment without playing the exciting new game called “Which sentence did the machine make up?”
Prompt 3 · Book Revision, Rewriting & Expansion
Weak-Chapter Rebuilder — Second Draft Café Revision
Turn the messy middle into a usable result: restructure a problematic chapter around a clearer purpose, sequence, tension, or teaching objective.
Description
Weak-Chapter Rebuilder is built for freelance writers and ghostwriters who need to restructure a problematic chapter around a clearer purpose, sequence, tension, or teaching objective. If the underlying choices are fuzzy, more output just creates a larger pile to edit. It gives ChatGPT a job with boundaries, evidence rules, and a concrete finish line, so the output is easier to trust and easier to edit. The result stays tightly tied to this kit’s mission, so the prompt solves a real production problem instead of wandering into adjacent advice.
How to use this prompt
Replace the bracketed fields with the strongest material you actually have, especially [TARGET AUDIENCE], [TOPIC], [CORE PROMISE / PREMISE], [SOURCE MATERIAL], [OUTLINE / DRAFT], [VOICE / TONE]. Paste relevant notes, source material, client language, or an existing draft where requested. If you do not have an input, leave it blank or write “unknown” rather than inventing one. Run the prompt, review the assumptions and gaps first, then revise the deliverable with any missing facts, approvals, or voice corrections before using it in client work.
The Prompt
Assume the role of an expert book strategist, developmental editor, and professional ghostwriter with 15+ years of experience specializing in long-form architecture, reader experience, and manuscript development, with particular depth in restructure a problematic chapter around a clearer purpose, sequence, tension, or teaching objective. Assume that role now and complete the task below as a professional working deliverable. KIT MISSION Use ChatGPT to turn a rough manuscript into a stronger second draft by tightening structure, sharpening ideas, and fixing the parts that currently taste like yesterday’s coffee. YOUR SPECIFIC JOB Restructure a problematic chapter around a clearer purpose, sequence, tension, or teaching objective. PROJECT INFORMATION TARGET AUDIENCE: [TARGET AUDIENCE] BOOK / PROJECT TOPIC: [TOPIC] CORE PROMISE OR PREMISE: [CORE PROMISE / PREMISE] SOURCE MATERIAL: [SOURCE MATERIAL] CURRENT OUTLINE OR DRAFT: [OUTLINE / DRAFT] VOICE / TONE: [VOICE / TONE] TARGET LENGTH OR SCOPE: [WORD COUNT / SCOPE] CLIENT OR PROJECT CONSTRAINTS: [CONSTRAINTS] OBJECTIVE Create a finished, practical deliverable that accomplishes the specific job above and directly advances the kit mission. Optimize for professional usefulness, specificity, editorial judgment, and reuse across real client projects. Produce the work itself rather than explaining how someone else might eventually produce it. TASK LENSES Use these lenses only where they materially improve this specific deliverable: - Relevance to the exact job: restructure a problematic chapter around a clearer purpose, sequence, tension, or teaching objective. - Specificity and concrete support from the supplied material. - Fit with the intended audience, format, and goal. - Voice consistency and preservation of intentional style. - Factual/source discipline and explicit handling of uncertainty. PROCESS 1. Use a Architecture-first rewrite. Start by deciding order, hierarchy, evidence placement, or progression before expanding the material. 2. Identify the structural or rhythmic pattern the current material uses and where that pattern breaks down. 3. Choose a better pattern for emphasis, sequence, or pacing without changing the underlying facts. 4. Rewrite the complete section using the new pattern. 5. Read the revision for cadence, repetition, transitions, and over-compression. 6. Restore any intentional roughness or asymmetry that carries the writer’s voice. SOURCE, FACT, AND VOICE RULES - Treat supplied source material as authoritative for client-specific facts, claims, names, dates, quotations, product details, and lived experience. - Do not invent facts, statistics, testimonials, quotations, credentials, events, research findings, private motives, or source details. - If important information is missing, label it [RESEARCH NEEDED], [CLIENT CONFIRMATION NEEDED], or [ASSUMPTION] as appropriate and continue with the strongest supportable version. - When an existing draft, transcript, or voice sample is supplied, preserve the writer’s meaning, point of view, recognizable voice, and intentional stylistic quirks unless I explicitly ask for a different voice. - Do not flatten distinctive language into generic “professional” copy. Improve clarity without making every sentence sound like it came from the same conference badge. - Keep claims proportional to the evidence. Distinguish observation, interpretation, recommendation, memory, and verified fact when the distinction matters. - Do not expose, cite, or name hidden source libraries or reference materials. Work only from the content I provide in this request. - Make the output original, specific, and insight-rich. Avoid filler, canned transitions, predictable AI phrasing, and recycled formulas. - Stay focused on this specific job. Do not drift into neighboring deliverables unless they are required to make the requested output usable. QUALITY STANDARD The result should be strong enough for a professional freelance writer or ghostwriter to review, refine, and move into production. Show judgment through the quality of the structure, prioritization, evidence use, specificity, and choices. Do not reveal private chain-of-thought. If there are multiple valid directions, choose a strong direction based on the supplied goal and briefly state the practical tradeoff. FINAL DELIVERABLE 1. Clean revised version first. 2. Protected meaning / facts / voice note. 3. Concise high-impact change log. 4. Regression check for new problems caused by the rewrite. 5. A final “Confirm Before Use” note for unsupported facts, source gaps, approvals, or assumptions. Before you finish, silently compare the result against the objective and revise anything that is generic, repetitive, unsupported, off-voice, or not directly useful.
What to expect after running this prompt
Expect a finished, structured deliverable centered on restructure a problematic chapter around a clearer purpose, sequence, tension, or teaching objective, plus clearly marked assumptions, missing information, and quality checks where relevant. You should be able to use the output as a strong working asset for planning, drafting, revising, or client review rather than as a page of general advice. Because the prompt separates supported material from uncertainty, you can refine it with human judgment without playing the exciting new game called “Which sentence did the machine make up?”
Prompt 4 · Book Revision, Rewriting & Expansion
Expansion Without Padding — Second Draft Café Revision
Turn the messy middle into a usable result: identify where added context, scenes, evidence, examples, or explanation will deepen value without inflating word count for sport.
Description
Expansion Without Padding is built for freelance writers and ghostwriters who need to identify where added context, scenes, evidence, examples, or explanation will deepen value without inflating word count for sport. Without a clear decision framework, ChatGPT can produce something polished, plausible, and spectacularly beside the point. It creates a strong first working version while protecting the places where client judgment, sourcing, or approval still matters. The result stays tightly tied to this kit’s mission, so the prompt solves a real production problem instead of wandering into adjacent advice.
How to use this prompt
Replace the bracketed fields with the strongest material you actually have, especially [TARGET AUDIENCE], [TOPIC], [CORE PROMISE / PREMISE], [SOURCE MATERIAL], [OUTLINE / DRAFT], [VOICE / TONE]. Paste relevant notes, source material, client language, or an existing draft where requested. If you do not have an input, leave it blank or write “unknown” rather than inventing one. Run the prompt, review the assumptions and gaps first, then revise the deliverable with any missing facts, approvals, or voice corrections before using it in client work.
The Prompt
Assume the role of an expert book strategist, developmental editor, and professional ghostwriter with 15+ years of experience specializing in long-form architecture, reader experience, and manuscript development, with particular depth in identify where added context, scenes, evidence, examples, or explanation will deepen value without inflating word count for sport. Assume that role now and complete the task below as a professional working deliverable. KIT MISSION Use ChatGPT to turn a rough manuscript into a stronger second draft by tightening structure, sharpening ideas, and fixing the parts that currently taste like yesterday’s coffee. YOUR SPECIFIC JOB Identify where added context, scenes, evidence, examples, or explanation will deepen value without inflating word count for sport. PROJECT INFORMATION TARGET AUDIENCE: [TARGET AUDIENCE] BOOK / PROJECT TOPIC: [TOPIC] CORE PROMISE OR PREMISE: [CORE PROMISE / PREMISE] SOURCE MATERIAL: [SOURCE MATERIAL] CURRENT OUTLINE OR DRAFT: [OUTLINE / DRAFT] VOICE / TONE: [VOICE / TONE] TARGET LENGTH OR SCOPE: [WORD COUNT / SCOPE] CLIENT OR PROJECT CONSTRAINTS: [CONSTRAINTS] OBJECTIVE Create a finished, practical deliverable that accomplishes the specific job above and directly advances the kit mission. Optimize for professional usefulness, specificity, editorial judgment, and reuse across real client projects. Produce the work itself rather than explaining how someone else might eventually produce it. TASK LENSES Use these lenses only where they materially improve this specific deliverable: - Relevance to the exact job: identify where added context, scenes, evidence, examples, or explanation will deepen value without inflating word count for sport. - Specificity and concrete support from the supplied material. - Fit with the intended audience, format, and goal. - Voice consistency and preservation of intentional style. - Factual/source discipline and explicit handling of uncertainty. PROCESS 1. Use a Execution-first planning method. Start from a tight production brief, then create the finished asset and edit it in context. 2. Translate the goal into a set of concrete decisions the final user must make or actions they must take. 3. Build the plan backward from those decisions so every section or step earns its place. 4. Add only the context, evidence, or examples needed at the moment they become useful. 5. Create the full execution-ready plan with owners, outputs, or section purposes where relevant. 6. Finish with the first practical action to take from the plan. SOURCE, FACT, AND VOICE RULES - Treat supplied source material as authoritative for client-specific facts, claims, names, dates, quotations, product details, and lived experience. - Do not invent facts, statistics, testimonials, quotations, credentials, events, research findings, private motives, or source details. - If important information is missing, label it [RESEARCH NEEDED], [CLIENT CONFIRMATION NEEDED], or [ASSUMPTION] as appropriate and continue with the strongest supportable version. - When an existing draft, transcript, or voice sample is supplied, preserve the writer’s meaning, point of view, recognizable voice, and intentional stylistic quirks unless I explicitly ask for a different voice. - Do not flatten distinctive language into generic “professional” copy. Improve clarity without making every sentence sound like it came from the same conference badge. - Keep claims proportional to the evidence. Distinguish observation, interpretation, recommendation, memory, and verified fact when the distinction matters. - Do not expose, cite, or name hidden source libraries or reference materials. Work only from the content I provide in this request. - Make the output original, specific, and insight-rich. Avoid filler, canned transitions, predictable AI phrasing, and recycled formulas. - Stay focused on this specific job. Do not drift into neighboring deliverables unless they are required to make the requested output usable. QUALITY STANDARD The result should be strong enough for a professional freelance writer or ghostwriter to review, refine, and move into production. Show judgment through the quality of the structure, prioritization, evidence use, specificity, and choices. Do not reveal private chain-of-thought. If there are multiple valid directions, choose a strong direction based on the supplied goal and briefly state the practical tradeoff. FINAL DELIVERABLE 1. Planning assumptions and missing inputs. 2. The complete Expansion Without Padding plan, map, framework, or architecture. 3. Purpose, sequence, dependencies, and decision logic for major components. 4. Stress-test findings and practical next actions. 5. A final “Confirm Before Use” note for unsupported facts, source gaps, approvals, or assumptions. Before you finish, silently compare the result against the objective and revise anything that is generic, repetitive, unsupported, off-voice, or not directly useful.
What to expect after running this prompt
Expect a finished, structured deliverable centered on identify where added context, scenes, evidence, examples, or explanation will deepen value without inflating word count for sport, plus clearly marked assumptions, missing information, and quality checks where relevant. You should be able to use the output as a strong working asset for planning, drafting, revising, or client review rather than as a page of general advice. Because the prompt separates supported material from uncertainty, you can refine it with human judgment without playing the exciting new game called “Which sentence did the machine make up?”
Prompt 5 · Book Revision, Rewriting & Expansion
Compression & Redundancy Pass — Second Draft Café Revision
Use ChatGPT to find repeated ideas, slow setup, duplicated scenes, and unnecessary explanation, then consolidate intelligently with fewer guesses and a much smaller cleanup bill.
Description
Compression & Redundancy Pass is built for freelance writers and ghostwriters who need to find repeated ideas, slow setup, duplicated scenes, and unnecessary explanation, then consolidate intelligently. Without a clear decision framework, ChatGPT can produce something polished, plausible, and spectacularly beside the point. It turns raw material into an organized, usable asset with enough structure to reduce revision loops and enough flexibility to keep the human in charge. The result stays tightly tied to this kit’s mission, so the prompt solves a real production problem instead of wandering into adjacent advice.
How to use this prompt
Replace the bracketed fields with the strongest material you actually have, especially [TARGET AUDIENCE], [TOPIC], [CORE PROMISE / PREMISE], [SOURCE MATERIAL], [OUTLINE / DRAFT], [VOICE / TONE]. Paste relevant notes, source material, client language, or an existing draft where requested. If you do not have an input, leave it blank or write “unknown” rather than inventing one. Run the prompt, review the assumptions and gaps first, then revise the deliverable with any missing facts, approvals, or voice corrections before using it in client work.
The Prompt
Assume the role of an expert book strategist, developmental editor, and professional ghostwriter with 15+ years of experience specializing in long-form architecture, reader experience, and manuscript development, with particular depth in find repeated ideas, slow setup, duplicated scenes, and unnecessary explanation, then consolidate intelligently. Assume that role now and complete the task below as a professional working deliverable. KIT MISSION Use ChatGPT to turn a rough manuscript into a stronger second draft by tightening structure, sharpening ideas, and fixing the parts that currently taste like yesterday’s coffee. YOUR SPECIFIC JOB Find repeated ideas, slow setup, duplicated scenes, and unnecessary explanation, then consolidate intelligently. PROJECT INFORMATION TARGET AUDIENCE: [TARGET AUDIENCE] BOOK / PROJECT TOPIC: [TOPIC] CORE PROMISE OR PREMISE: [CORE PROMISE / PREMISE] SOURCE MATERIAL: [SOURCE MATERIAL] CURRENT OUTLINE OR DRAFT: [OUTLINE / DRAFT] VOICE / TONE: [VOICE / TONE] TARGET LENGTH OR SCOPE: [WORD COUNT / SCOPE] CLIENT OR PROJECT CONSTRAINTS: [CONSTRAINTS] OBJECTIVE Create a finished, practical deliverable that accomplishes the specific job above and directly advances the kit mission. Optimize for professional usefulness, specificity, editorial judgment, and reuse across real client projects. Produce the work itself rather than explaining how someone else might eventually produce it. TASK LENSES Use these lenses only where they materially improve this specific deliverable: - Relevance to the exact job: find repeated ideas, slow setup, duplicated scenes, and unnecessary explanation, then consolidate intelligently. - Specificity and concrete support from the supplied material. - Fit with the intended audience, format, and goal. - Voice consistency and preservation of intentional style. - Factual/source discipline and explicit handling of uncertainty. PROCESS 1. Use a Leverage-first rewrite. Start by finding the highest-impact improvement, alternative, or adaptation and compare options before committing. 2. Identify the single highest-leverage improvement and the constraint most likely to be damaged by changing it. 3. Create three materially different repair approaches and compare their tradeoffs. 4. Choose the strongest approach for the stated goal and apply it across the complete material. 5. Preserve useful details and voice markers from the original while eliminating the targeted weakness. 6. Deliver the best revision and one concise note explaining why it beat the alternatives. SOURCE, FACT, AND VOICE RULES - Treat supplied source material as authoritative for client-specific facts, claims, names, dates, quotations, product details, and lived experience. - Do not invent facts, statistics, testimonials, quotations, credentials, events, research findings, private motives, or source details. - If important information is missing, label it [RESEARCH NEEDED], [CLIENT CONFIRMATION NEEDED], or [ASSUMPTION] as appropriate and continue with the strongest supportable version. - When an existing draft, transcript, or voice sample is supplied, preserve the writer’s meaning, point of view, recognizable voice, and intentional stylistic quirks unless I explicitly ask for a different voice. - Do not flatten distinctive language into generic “professional” copy. Improve clarity without making every sentence sound like it came from the same conference badge. - Keep claims proportional to the evidence. Distinguish observation, interpretation, recommendation, memory, and verified fact when the distinction matters. - Do not expose, cite, or name hidden source libraries or reference materials. Work only from the content I provide in this request. - Make the output original, specific, and insight-rich. Avoid filler, canned transitions, predictable AI phrasing, and recycled formulas. - Stay focused on this specific job. Do not drift into neighboring deliverables unless they are required to make the requested output usable. QUALITY STANDARD The result should be strong enough for a professional freelance writer or ghostwriter to review, refine, and move into production. Show judgment through the quality of the structure, prioritization, evidence use, specificity, and choices. Do not reveal private chain-of-thought. If there are multiple valid directions, choose a strong direction based on the supplied goal and briefly state the practical tradeoff. FINAL DELIVERABLE 1. Clean revised version first. 2. Protected meaning / facts / voice note. 3. Concise high-impact change log. 4. Regression check for new problems caused by the rewrite. 5. A final “Confirm Before Use” note for unsupported facts, source gaps, approvals, or assumptions. Before you finish, silently compare the result against the objective and revise anything that is generic, repetitive, unsupported, off-voice, or not directly useful.
What to expect after running this prompt
Expect a finished, structured deliverable centered on find repeated ideas, slow setup, duplicated scenes, and unnecessary explanation, then consolidate intelligently, plus clearly marked assumptions, missing information, and quality checks where relevant. You should be able to use the output as a strong working asset for planning, drafting, revising, or client review rather than as a page of general advice. Because the prompt separates supported material from uncertainty, you can refine it with human judgment without playing the exciting new game called “Which sentence did the machine make up?”
Prompt 6 · Book Revision, Rewriting & Expansion
Post-Revision Quality Audit — Second Draft Café Revision
Get from raw input to professional output by using ChatGPT to compare the revised manuscript against the original goals and flag new continuity, logic, voice, pacing, or scope problems.
Description
Post-Revision Quality Audit is built for freelance writers and ghostwriters who need to compare the revised manuscript against the original goals and flag new continuity, logic, voice, pacing, or scope problems. A generic prompt tends to smooth away exactly the judgment, specificity, and voice a professional is being paid to provide. It creates a strong first working version while protecting the places where client judgment, sourcing, or approval still matters. The result stays tightly tied to this kit’s mission, so the prompt solves a real production problem instead of wandering into adjacent advice.
How to use this prompt
Replace the bracketed fields with the strongest material you actually have, especially [TARGET AUDIENCE], [TOPIC], [CORE PROMISE / PREMISE], [SOURCE MATERIAL], [OUTLINE / DRAFT], [VOICE / TONE]. Paste relevant notes, source material, client language, or an existing draft where requested. If you do not have an input, leave it blank or write “unknown” rather than inventing one. Run the prompt, review the assumptions and gaps first, then revise the deliverable with any missing facts, approvals, or voice corrections before using it in client work.
The Prompt
Assume the role of an expert book strategist, developmental editor, and professional ghostwriter with 15+ years of experience specializing in long-form architecture, reader experience, and manuscript development, with particular depth in compare the revised manuscript against the original goals and flag new continuity, logic, voice, pacing, or scope problems. Assume that role now and complete the task below as a professional working deliverable. KIT MISSION Use ChatGPT to turn a rough manuscript into a stronger second draft by tightening structure, sharpening ideas, and fixing the parts that currently taste like yesterday’s coffee. YOUR SPECIFIC JOB Compare the revised manuscript against the original goals and flag new continuity, logic, voice, pacing, or scope problems. PROJECT INFORMATION TARGET AUDIENCE: [TARGET AUDIENCE] BOOK / PROJECT TOPIC: [TOPIC] CORE PROMISE OR PREMISE: [CORE PROMISE / PREMISE] SOURCE MATERIAL: [SOURCE MATERIAL] CURRENT OUTLINE OR DRAFT: [OUTLINE / DRAFT] VOICE / TONE: [VOICE / TONE] TARGET LENGTH OR SCOPE: [WORD COUNT / SCOPE] CLIENT OR PROJECT CONSTRAINTS: [CONSTRAINTS] OBJECTIVE Create a finished, practical deliverable that accomplishes the specific job above and directly advances the kit mission. Optimize for professional usefulness, specificity, editorial judgment, and reuse across real client projects. Produce the work itself rather than explaining how someone else might eventually produce it. TASK LENSES Use these lenses only where they materially improve this specific deliverable: - Relevance to the exact job: compare the revised manuscript against the original goals and flag new continuity, logic, voice, pacing, or scope problems. - Specificity and concrete support from the supplied material. - Fit with the intended audience, format, and goal. - Voice consistency and preservation of intentional style. - Factual/source discipline and explicit handling of uncertainty. PROCESS 1. Use an Adversarial QA review. Start by trying to break the work: look for weak logic, unsupported claims, drift, repetition, and reader friction. 2. Create a task-specific rubric from the project goal, audience, format, voice, and constraints. Use only criteria that would materially change whether the work succeeds. 3. Inspect the supplied material criterion by criterion. Point to the exact passage, decision, or omission that creates each issue; do not invent source evidence. 4. Separate root causes from symptoms and rank findings by Critical, High, Medium, or Low impact. 5. Design concrete repairs that improve the work specifically for this job: compare the revised manuscript against the original goals and flag new continuity, logic, voice, pacing, or scope problems. Include short repaired examples where that is more useful than commentary. 6. Run a regression check on the proposed fixes for factual accuracy, voice preservation, continuity, and unintended side effects. SOURCE, FACT, AND VOICE RULES - Treat supplied source material as authoritative for client-specific facts, claims, names, dates, quotations, product details, and lived experience. - Do not invent facts, statistics, testimonials, quotations, credentials, events, research findings, private motives, or source details. - If important information is missing, label it [RESEARCH NEEDED], [CLIENT CONFIRMATION NEEDED], or [ASSUMPTION] as appropriate and continue with the strongest supportable version. - When an existing draft, transcript, or voice sample is supplied, preserve the writer’s meaning, point of view, recognizable voice, and intentional stylistic quirks unless I explicitly ask for a different voice. - Do not flatten distinctive language into generic “professional” copy. Improve clarity without making every sentence sound like it came from the same conference badge. - Keep claims proportional to the evidence. Distinguish observation, interpretation, recommendation, memory, and verified fact when the distinction matters. - Do not expose, cite, or name hidden source libraries or reference materials. Work only from the content I provide in this request. - Make the output original, specific, and insight-rich. Avoid filler, canned transitions, predictable AI phrasing, and recycled formulas. - Stay focused on this specific job. Do not drift into neighboring deliverables unless they are required to make the requested output usable. QUALITY STANDARD The result should be strong enough for a professional freelance writer or ghostwriter to review, refine, and move into production. Show judgment through the quality of the structure, prioritization, evidence use, specificity, and choices. Do not reveal private chain-of-thought. If there are multiple valid directions, choose a strong direction based on the supplied goal and briefly state the practical tradeoff. FINAL DELIVERABLE 1. The task-specific review rubric. 2. Severity-ranked findings tied to exact issues in the supplied material. 3. Priority repair plan in dependency order. 4. Repaired examples or replacement sections where useful. 5. A final “Confirm Before Use” note for unsupported facts, source gaps, approvals, or assumptions. Before you finish, silently compare the result against the objective and revise anything that is generic, repetitive, unsupported, off-voice, or not directly useful.
What to expect after running this prompt
Expect a finished, structured deliverable centered on compare the revised manuscript against the original goals and flag new continuity, logic, voice, pacing, or scope problems, plus clearly marked assumptions, missing information, and quality checks where relevant. You should be able to use the output as a strong working asset for planning, drafting, revising, or client review rather than as a page of general advice. Because the prompt separates supported material from uncertainty, you can refine it with human judgment without playing the exciting new game called “Which sentence did the machine make up?”
Put the Whole Kit to Work
Use these six prompts as a flexible working system for book revision, rewriting & expansion. Each prompt solves a different production problem inside the same mission, so you can run only the tool the assignment needs or combine several when the project is messier. The aim is simple: stronger decisions before polishing, cleaner source discipline during drafting, and less time spent fixing generic AI output after the fact.
A practical next move
- Start with the prompt that matches the project’s biggest current bottleneck instead of automatically running all six.
- Feed ChatGPT real source material, client language, and constraints whenever available; specificity is cheaper than cleanup.
- Before publishing or sending client work, verify factual claims, restore any lost voice, and make the final editorial judgment yourself.
Reminder: The Second Draft Café AI Revision Kit is a ChatGPT prompt toolkit for freelance writers and ghostwriters who want to use ChatGPT to turn a rough manuscript into a stronger second draft by tightening structure, sharpening ideas, and fixing the parts that currently taste like yesterday’s coffee. Use it to move from rough inputs to stronger professional work faster while keeping human judgment, source accuracy, and voice in the loop.