agentleFS
renky1025renky1025/agent-skills/novel-writing/SKILL.md

novel-writing

Skill11 starsChanged 50 days agoMITView on GitHub

Build fiction from four pillars — LOCK (Lead/Objective/Conflict/Knockout), Three-Act + Two Doors, Dialogue & Conflict, and Character Arc — and produce story outlines, casts, scene plans, and prose. Use when writing fiction, developing novel plots, creating characters, designing story structure, crafting dialogue, or planning character arcs. Triggers when the user asks about story outlining, scene writing, conflict design, or any creative writing task involving narrative structure. 触发词:写小说、故事大纲、情节设计、人物设定、人物弧光、场景写作、冲突设计、对白、novel outline、story structure、character arc、scene writing、dialogue.

6 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Before reporting the work done:

    line 83

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Novel Writing Agent

LOCK system + three-act structure + character arcs — applied to fiction creation.

Purpose

Provide a systematic framework for novel creation built on four pillars: LOCK (Lead, Objective, Conflict, Knockout), Three-Act + Two Doors structure, Dialogue & Conflict, and Character Arc design. The skill constructs cohesive stories rather than stitching isolated scenes.

When to Use

  • Planning or outlining a new story.
  • Developing protagonists, antagonists, or supporting cast.
  • Structuring plot across acts.
  • Writing dialogue with tension.
  • Designing character growth trajectories.
  • Identifying and fixing story problems (flat characters, sagging middle, weak climax).
  • Triggers: "写小说", "故事大纲", "情节设计", "人物弧光", "对白", "story outlining", "scene writing", "conflict design", "character arc", "novel plot".

When NOT to Use

  • User needs a fine-grained layer-by-layer Chinese-first workflow (one sentence -> paragraph -> page -> draft) with built-in anti-AI-tone and anti-cliché checks -> hand off to snowflake-novel-writer.
  • User wants AI-tone removal on existing prose rather than structure work -> hand off to de-ai-writing.
  • User wants a marketing / article piece, not fiction -> hand off to wechat-article-writer.
  • User only asks for a one-line logline with no structural work -> answer directly.

Workflow

  1. Spark — Extract story seeds from the user's premise. Ask what emotions, questions, or images drive the idea.
  2. Structure — Propose 2-3 structural frameworks with tradeoffs. User selects and modifies. Lock in act breakpoints and Door events.
  3. Cast — Build characters top-down: Lead first, then opposition, then supporting. Each gets a LOCK mini-profile (who they are, what they want, what blocks them, how they land).
  4. Scenes — Draft scene by scene. After each major beat, check alignment with structure. Offer branches when the user hits a wall.
  5. Dialogue — Write dialogue with embedded conflict. Evaluate against speaker profile — does this person talk like this?
  6. Polish — Focus passes on language (precision, rhythm), logic (cause and effect), and consistency (character voice, world rules). Show before/after comparisons.

Decision Rules

Install the right pillar first based on the symptom:

TaskStart WithWatch For
New storyLOCK profile for LeadGeneric objectives — make them specific and personal
Sagging middleDoor 2 placementLack of escalation — each beat must raise stakes
Flat characterArc Start -> End contrastChange too small or too sudden
Static dialogueConflict type mismatchCharacters agreeing too much — inject opposition
Weak climaxKnockout testCan the reader guess the outcome? If yes, redesign
  • Need the LOCK pillars in detail -> read references/lock-system.md.
  • Need act/breakpoint definitions -> read references/three-act-structure.md.
  • Need dialogue conflict types and principles -> read references/dialogue-and-conflict.md.
  • Need the arc model -> read references/character-arc.md.
  • Draft has a specific defect -> read references/common-mistakes.md.

Constraints

  • Conflict must be logical for the characters involved and must escalate organically, not by authorial force.
  • Subtext beats statement: prefer hesitation, truncated sentences, and misdirection to exposition.
  • Vocabulary and syntax must match each speaker's background.
  • A character's transformation must be driven by a decision, not by an event.
  • Propose 2-3 structural options and let the user choose; never silently impose one framework.

Verification

Before reporting the work done:

  • Lead has a specific, personal Objective (not a generic one).
  • Door 1 and Door 2 are each identifiable in the outline.
  • Every scene answers "what does the Lead want now, and what blocks it?"
  • Dialogue lines pass the blind-voice test (speaker is guessable).
  • The arc's turning point is a choice, not an accident.
  • The climax survives the Knockout test (outcome is not guessable).

Output

  • Outcome: A story package — locked structure with Door events, LOCK profiles for the cast, a scene plan, and drafted/revised prose.
  • Done when: the requested artifact (outline, cast sheet, scene, or polished draft) is delivered and passes the Verification checklist.
  • Evidence: named Door 1 / Door 2 events; per-character LOCK mini-profiles; dialogue that carries embedded conflict; before/after comparison from the Polish pass.

References

Load only when needed:

  • references/lock-system.md — read when establishing or auditing the story's four pillars.
  • references/three-act-structure.md — read when setting act breakpoints or fixing pacing.
  • references/dialogue-and-conflict.md — read when writing or repairing dialogue.
  • references/character-arc.md — read when designing or debugging a character transformation.
  • references/common-mistakes.md — read during diagnosis or the Polish pass.

Related Skills

  • snowflake-novel-writer — Chinese-first, ten-step layer-by-layer workflow with anti-cliché and anti-AI-tone checks; use after LOCK has settled the skeleton.
  • de-ai-writing — the single source of truth for AI-tone removal.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.