A One-Line Premise Is Not a Story Engine
Start by separating two things:
- A premise answers: who is caught in what dramatic situation?
- A story engine answers: what pressure can return in different forms, what method the protagonist uses, how opposition learns, what changed state the next round must carry forward, and what new development could end the loop or revise that state?
This guide uses an original, everyday premise:
Zhou Yao, the acting coordinator at a neighborhood service center, wants to keep every resident's request from falling through while the director is away.
That line gives us a character, a setting, and a desire, but it cannot yet predict the next conflict. If every cycle merely introduces “another more difficult resident” and restores the people and rules to their starting positions without a cause, the story is repeating incidents rather than running an engine. A sustainable version makes Zhou Yao's useful method create a new cost every time. Each allocation decision changes who has priority, who controls evidence, who will cooperate, and whether the community trusts the rules. Later rounds must inherit those states unless a new choice and result causally change them.
Developing the idea therefore means finding a loop that produces choice, countermove, changed state, and new pressure, not listing more events. The loop may start repeatedly, but earlier state cannot reset without a cause. The rounds below are a stress test, not a prescribed episode count, turn count, or platform formula.
Write a Compact Creative Brief First
A creative brief records only decisions that constrain later choices. Mark unknowns as “to be confirmed” instead of filling the page with lore or guessing on the creator's behalf. Start with these fields:
| Field | Example |
|---|---|
| Original idea | Zhou Yao runs the neighborhood service center temporarily and wants every request to find an answer |
| Repeatable audience reward | A seemingly fair rule solves the immediate problem, then exposes a new cost in interests or relationships |
| Thematic tension | Equal treatment versus care for a particular circumstance |
| Tone and boundaries | Realistic and humane without avoiding material conflict; people seeking help are neither jokes nor villains |
| Focus and pursuit | Zhou Yao wants to preserve the center's service promise and prove she can coordinate on her own |
| Costly resistance | Volunteer hours and service slots are limited; residents hold different evidence and public influence, and any exception may become precedent |
| Distinctive contradiction | The more openly Zhou Yao states a rule, the easier it is to challenge through that rule; the more she makes private exceptions, the less credible the system becomes |
| Terminal change | The old engine transforms once allocation becomes a responsibility residents share, or once Zhou Yao stops acting from “I must satisfy everyone” |
Add only the project's real constraints: primary language, viewing context, feasible cast and locations, unaffordable production elements, and any ending promise or prohibited outcome. This stage confirms direction. It does not write dialogue, stage scenes, design an opening episode, or map episode beats. If a field cannot affect a character choice or a content boundary, remove it for now rather than turning the brief into an encyclopedia.
These tables are editorial excerpts for comparison, not complete copies of native files. The skill drafts creative-brief.md and story-engine.md for the relevant entry point and can deepen an already chosen direction. Complete the source, character, and acceptance information the actual project needs; do not force every example field on every work.
Compare Two Mechanisms, Not Two Labels
Candidate directions must change the protagonist's method, the source of opposition, the cost, or the way conflict regenerates. Both options below preserve the neighborhood-service premise, but they run on different mechanisms:
| Item | Two directions |
|---|---|
| Protagonist's method | Direction A: public allocation rules: Zhou Yao publishes priority criteria, required evidence, and reasons for changes, then allocates scarce services through an explainable order; Direction B: a mutual-aid credit network: Instead of allocating alone, Zhou Yao matches people who need help with people who can offer time or skills, allowing assistance to create an optional reciprocal commitment |
| Opposition mechanism | Direction A: public allocation rules: Residents add evidence, reframe urgency, or organize challenges to exceptions; every ruling creates a precedent others can cite; Direction B: a mutual-aid credit network: Participants hide capabilities, reject social debt, transfer commitments, or form cliques; debt and trust continually reshape which matches are possible |
| Central cost | Direction A: public allocation rules: Zhou Yao absorbs the reputational and relational cost of public judgment; the clearer the rule, the more precisely its gaps can be attacked; Direction B: a mutual-aid credit network: Zhou Yao learns more private circumstances; a failed match damages trust on both sides and may turn mutual aid into moral pressure |
| Repeatable reward | Direction A: public allocation rules: Viewers see one fairness decision redistribute resources, voice, and authority to interpret the rules; Direction B: a mutual-aid credit network: Viewers see one exchange create a new alliance, obligation, or refusal |
| Terminal condition | Direction A: public allocation rules: Shared governance replaces unilateral allocation, or Zhou Yao gives up the role of sole judge; Direction B: a mutual-aid credit network: Mutual aid no longer depends on Zhou Yao's guarantee, or the community rejects trading services through relational debt |
Direction B is not the same queue renamed as a mutual-aid list. It changes Zhou Yao from judge to matchmaker, moves opposition from rule disputes to debt and trust, and shifts the cost of failure from public legitimacy to private relationships and privacy.
This example selects Direction A because it carries the tension between equal treatment and particular care most directly, while allowing changes in power, evidence, and alliances to remain visible in one ordinary service center. Direction B stays unselected and should not leak into the accepted engine. Combining them would require the creator to reconsider the costs and terminal condition.
Compress the Chosen Direction into an Engine Card
An engine card is not a plot synopsis. It records a causal loop that can predict the next choice:
Source of pressure: A new request competes with an existing commitment for the same volunteer hours or service slots, while the available information is always incomplete.
Protagonist's usual method: Zhou Yao openly ranks urgency, number of people affected, and available alternatives, then explains each adjustment.
Why it works: Scarce resources are no longer allocated through personal connections or whoever speaks loudest; decisions can be examined.
How it creates a new problem: Residents gather evidence around the criteria, form alliances, or identify exceptions. One accommodation becomes leverage in the next dispute.
Opposition and adaptation: People seeking help, volunteers, and neighborhood committee members protect their time, reputation, and constituencies. They change what information they submit, withdraw cooperation, or demand shared review.
Result of one cycle: One request receives an answer, but at least one state changes: remaining resources, authority to interpret the rules, distribution of evidence, or relational debt. Later rounds must carry it forward unless a new causal result changes it again.
Next pressure: The new precedent qualifies another need or gives a former collaborator a reason to refuse.
Why the protagonist cannot simply leave: Zhou Yao has made commitments on behalf of the center. Walking away would leave unfinished requests and distrust for her colleagues to absorb, while breaking her self-image as a reliable person.
Terminal transformation: Residents and volunteers begin sharing responsibility for standards and exceptions, or Zhou Yao accepts that reliability does not mean satisfying everyone and stops monopolizing judgment and blame.
Content and production boundaries: Conflict comes from scarce services and relational choices, not exaggerated villains, sudden crime, or lavish locations. The action can remain within an ordinary service center, its public notices, and existing neighborhood-service processes.
The causal link between the local result and the next pressure is the crucial part. If the new pressure could be replaced with any unrelated problem, the card can generate incidents in the same genre, but not a single evolving story.
Stress-Test the Engine for Real State Change
The three rounds below test the engine. They are not a three-episode outline or mandatory plot. Each round starts from the state left by the previous one:
| Round | Carried results |
|---|---|
| Round one: resources | Trigger and choice: A medical escort and temporary childcare compete for the last volunteer hours. Zhou Yao applies “no available alternative comes first” and assigns the escort; Countermove and local result: The childcare applicant shows that the supposed alternative would miss work. The escort is completed, but the volunteer pool is exhausted; State carried forward: Resource change: no hours remain that day. Relationship change: rejected applicants organize around a challenge to the meaning of “available alternative” |
| Round two: knowledge and interpretive authority | Trigger and choice: The group submits evidence that one childcare assignment would keep several night-shift parents at work. Zhou Yao revises the meaning of “number affected” and opens a public appeal-and-review channel; Countermove and local result: The earlier beneficiary worries that a completed decision could be reversed retroactively, while a committee member demands that every appeal record its basis and status. The channel begins accepting cases; State carried forward: Knowledge change: “number affected” gains a new verifiable meaning. Power change: Zhou Yao loses room to interpret the rule alone, and the committee gains a review role. Process change: later disputes must enter a trackable pending queue |
| Round three: appeals and shared responsibility | Trigger and choice: The new channel receives two related matters: the childcare applicant requests the next available hours under the revised criteria; the earlier escort recipient asks that the completed service not be relabeled unjustified without new evidence. They enter one review agenda but do not compete for the same hours. Today's pool remains empty, so the future childcare request stays queued; Countermove and local result: If Zhou Yao rules alone again, public review becomes another form of personal discretion. The organized applicants will join the review only if they can challenge unsupported exceptions; the earlier recipient asks reviewers to keep the completed service record separate from future allocation. The committee sets a rule: absent new evidence, completed service is not rewritten retroactively, while future requests follow the revised criteria; State carried forward: Resource state carried forward: no hours remain that day, so the future childcare request stays queued; the completed escort remains a valid past record. Relationship change: challengers and an existing beneficiary enter shared review. Power change: responsibility for review and future allocation is shared. Character change: Zhou Yao must replace solitary rescue with shared responsibility |
Now ask four questions:
- Must each later round inherit a resource, knowledge, power, or relationship state from the previous one unless a new choice and result causally revise it?
- Does the opposing force learn from results and change its method rather than merely becoming louder?
- Does the protagonist's method solve a problem while creating a cost she genuinely cares about?
- Can you name a change in character, fact, or relationship that would naturally end or transform the old loop?
If not, revise the direction or engine card. Do not hide the break by adding an unrelated villain, secret identity, or more lore. Once the test passes and the creator accepts the brief and engine, they can guide downstream work. Use the short-drama episode outline only when you are ready to assign entry states, local results, and handoffs to episodes. Consult short-drama episode structure when testing how one episode completes a causal state change. Move to short-drama screenwriting only after those results are defined and you are ready for playable scenes. This guide stops here; it does not prewrite episodes or an opening hook.
FAQ
What is the difference between a logline and a story engine?
A logline summarizes a character, pursuit, and situation. A story engine also explains how pressure returns, how the protagonist and opposition adapt, what each cycle leaves for later cycles to inherit, and what would end the old loop. A logline may be compelling, but an engine lets you judge whether later conflicts arise from the same causal system.
Why compare two directions, and must I always write two?
When an idea has no mechanism yet, comparison exposes the real decision: the protagonist's method, the source of opposition, and the central cost. Two directions are enough to demonstrate that distinction here, but they are not a required count. If the creator has already chosen a mechanism, deepen it instead of inventing false alternatives.
Do the three stress-test rounds equal three episodes?
No. A round only tests whether one result causally creates the next pressure. A round may span episodes, or one episode may contain several rounds; decide that later according to the project. At this stage, test only whether later work inherits the state and changes it when a new cause appears.
When is the idea ready for an episode outline?
Begin episode mapping after the creator accepts the direction, the pursuit, costly resistance, repeatable reward, and terminal change agree with one another, and the engine produces consecutive states that later work either inherits or revises for a stated cause. Resolve any question that would still change the protagonist, opposition mechanism, or ending promise in the brief first.
Do it with the skill
In Claude Code, say: “Use /short-drama-develop to take this original short-drama idea only as far as a creative brief and story engine. Preserve my genre, tone, ending, and content boundaries; compare directions with genuinely different mechanisms; after I choose, write creative-brief.md and story-engine.md. Stress-test state changes across consecutive conflict rounds, but do not create an episode map, scenes, or dialogue yet.” In Codex, use $short-drama-develop. The skill explores directions for an idea-only project before building the brief and engine; the creator still chooses the direction.
Read the method: Development entry points and files · Mechanism loops and exhaustion