magnus919_agent-skills/writers-helper/references/workflow-and-planning.md

Workflow, Planning, and Research

The master map for any writing project, fiction or nonfiction. Use this reference when a project is being born: testing ideas, choosing structure, planning the work, and organizing research. It also covers the "goal math" that converts a deadline into a daily word target.

The writing lifecycle

Writing is a pipeline with seven stages: idea → plan → research → organize → draft → edit → evaluate. It is deliberately non-linear — drafting exposes research gaps, evaluation loops back into revision — but every piece passes through all seven eventually.

• The drafting and editing stages are separate states of mind, not just separate steps. Protect the draft (see drafting-and-process.md); do not let the editor touch the writer mid-draft.
• Every project needs a stated "done": a definition of completion and a deadline. Without both, the middle of any project becomes an open-ended grind.
• The agent's job is to name the current stage, keep the author's stated purpose visible, and move the project through the stages in order — while tolerating the zigzagging real writers do.

Finding and testing ideas

Idea-generation moves

• What-if and I-wonder chains. Take any premise and push it: "what if X" then "what if X and also Y". Invented logic may outrank literal truth — the factual "well, actually" kills the chain.
• Borrowed structure. Recast a known story in another genre, setting, or point of view. The structure is a scaffold; the new content is yours.
• Combinatorial lists. Build three lists — character types, conflicts, settings — and combine one from each. Two characters defined by a love-list vs. hate-list generates instant friction.
• Observation as fuel. Eavesdrop, people-watch, keep a notebook. The channels are three: observation, imagination, memory. Rotate them.
• Constraint prompts. "One emotion in five scenarios", "animate an object", "replay a decision with its alternate branch", "take the first line of a favorite book and make it your own".

The four-test screen for choosing an idea

1. Does it seize your own imagination? (You will live with it for months.)
2. Is the scope right? Card-outline it; can you finish it at the length you intend?
3. Would an audience care? (Required for publication-seeking work.)
4. Does it carry drama and conflict? A premise without conflict is a topic, not a story.

The one-line pitch and high-concept bar

• Logline: one to two sentences, under ~25 words: protagonist + goal + obstacle + stakes, written like poster tagline copy — short words, concrete imagery.
• Meets line: "Concept A meets Concept B" — two well-known touchstones that transmit tone and audience in one line. Use only successful, respected reference points.
• High-concept checklist: great title, fascinating subject, strong hook, broad audience appeal. If it can't be said in a couple of short sentences, it isn't high concept yet.
• Compression ladder: a writer must be able to state the story at every length: logline → meets line → one-page synopsis → outline/proposal → full draft. Never let anyone pitch at a length they haven't earned at shorter lengths.

Outlining: structure before prose

Two valid planning styles

• Tight (plotters): outline scenes in order before drafting. Strength: structure, completion, pacing. Weakness: characters and emotion can go flat; the outline can become a cage.
• Loose (pantsers): start with a premise, a protagonist, and a direction, and discover the story while writing. Strength: voice, surprise, emotional truth. Weakness: sagging middles, structural drift, unfinished drafts.
• The balanced move: each type compensates using its strength — plotters force themselves to develop the emotional line; pantsers force themselves to stake structure. Both end up with the same thing: a story that is plotted and felt.

Backward design

You cannot know the beginning until you write the end. When full outlining feels impossible:

1. Write the climax first: the scene where the transformed protagonist does something she could not do anywhere else in the story.
2. The climax dictates what skills the protagonist must be missing at the start and must relearn in the middle.
3. Stake the five load-bearing scenes: opening, Act I reversal, Act II reversal, climax, resolution. Fill the gaps later.

This is the antidote to perfectionism and to the "going-back-to-the-beginning" loop that stalls middle drafts.

Proportion map

• Beginning ≈ first quarter; middle ≈ half; end ≈ final quarter. The middle needs twice as many scenes as either end.
• Four energetic markers anchor the skeleton (full treatment in craft-and-structure.md):
  • ~25% — End of the Beginning: threshold crossed, no turning back
  • ~50% — Halfway Point: conscious recommitment to the journey
  • ~75% — Crisis: the figurative death of the old self, antagonist at full power
  • End — Climax: the transformed protagonist performs the action the old self could not

The outline as a revision instrument

Even writers who refuse to plan in advance must outline after writing. A scene-by-scene outline is the primary instrument for diagnosing pacing, chronology, subplot balance, and cause-and-effect gaps.

Worldbuilding (for speculative and historical fiction)

• Rules are the contract. Fantasy makes the impossible possible without explanation; science fiction makes it possible with one. In both, the author's entire job is to set internal rules and obey them. The purple fire in chapter 3 must still work in chapter 13, or its change must be shown growing from the story.
• The three-quarters rule. Build a world bible — map, magic/tech rules, religions, politics, species, geography — and expect most of it never to reach the page. The writing implies the mass that is cut.
• Translation policy. Render everything characters say in plain language; keep invented names, units, and idioms few, pronounceable, and only where they meaningfully change the story. No "Earth minutes."
• No infodumps. Deliver world facts through motivated dialogue and characters doing things, never exposition blocks. Give the reader only what the current scene needs.
• Anachronism scan. Check era-wrong words, objects, and food in historical fiction; dated technology in near-future SF.
• Don't demystify the wonder. Explaining the core magic as biology drains the awe; keep some mystery sacred.
• Depth of detail. Details develop like drafts: first-draft details are generic and clichéd; each rewrite makes them more specific and more universal.

Research discipline

• Write first, research later. Most pre-writing research never surfaces in the final draft. Draft with blank gaps marked, then hunt only the facts the draft actually needs. Research is a seducer that pulls writers away from the draft.
• Verification pass. Facts go stale — hours, prices, openings, titles. Verify or date-stamp claims. For nonfiction, details are the trust engine: concrete, verified specifics do more for credibility than any assertion of expertise.
• The gaps list. During drafting, maintain a running list of unknowns (names, dates, figures, descriptions) to resolve in a single post-draft research session.
• For memoir: interview relatives, dig through records and photos, revisit places, and treat honesty over completeness. A memoir is a careful construction of selected meaningful time — summary, exposition, and drama — not a whole-life autobiography.

Goal math: deadline to daily target

1. Pick a completion date.
2. Average the page counts of three books comparable to yours.
3. Multiply by ~250 words per page for the total word count.
4. Divide by the number of writing days remaining to get the daily target.
5. Schedule writing sessions in pen, and recalculate honestly when the calendar slips.

Example: three comparable novels average 320 pages → 80,000 words. Ninety writing days out → 889 words/day. Set the floor below that (~700) so days count as wins, and the ideal above it (~1,100).

Planning the nonfiction book: proposal before book

Nonfiction is sold as a proposal, not a finished manuscript: overview, market, competition, platform/promotion, chapter-by-chapter outline, two sample chapters (full structure in publishing-and-career.md and templates/book-proposal.md). Plan the book as the answer to a question the market is asking, with a specific audience you can reach.

Deliverables from this stage

• A one-line pitch and meets line that pass the high-concept bar
• A premise canvas (template: templates/premise-canvas.md)
• A story outline with the four markers staked (template: templates/story-outline.md)
• A character profile for the protagonist and antagonist (template: templates/character-profile.md)
• A worldbuilding questionnaire if the genre needs one (template: templates/worldbuilding-questionnaire.md)
• A daily word target and calendar
• A gaps list for post-draft research