fiction-workshop
Use when writing or editing novels, short stories, or any fiction manuscript. Trigger on: 'write fiction', 'edit my novel', 'developmental edit', 'line edit', 'character voice', 'plot hole', 'brainstorm', or fiction writing tasks.
How do I install this agent skill?
npx skills add https://github.com/shengchen1998/author-toolkit --skill fiction-workshopIs this agent skill safe to install?
- Gen Agent Trust Hubpass
The fiction-workshop skill provides a structured editorial workflow for fiction writing and editing. It includes comprehensive guidelines for story structure, character development, and prose refinement across multiple genres. No security risks or malicious patterns were detected.
- Socketpass
No alerts
- Snykpass
Risk: LOW · No issues
What does this agent skill do?
Fiction Workshop
Editorial workflow for collaborative fiction writing in three stages: Story Bible Building, Chapter Development, and Reader Testing.
Session Continuity
Novel projects span weeks or months, and agent sessions do not reliably preserve context. The applicable Story Bible or current-draft baseline is persistent creative state; .author-toolkit/project.json, when present, locates shared project artifacts.
- At session start: Read the baseline named by the caller, or
story-bible.mdin a standalone project. Skim recent files insessions/for unresolved threads. - At session end: Without
.author-toolkit/project.json, write a brief note atsessions/YYYY-MM-DD_topic-slug.mdsummarizing what was done, decisions made, backups created, and the stopping point. Two to five sentences is enough. In workflow work, return the stopping point to the caller for itsauthor-projectsession note; standalone work in a toolkit project leaves registered sessions unchanged. - When foundations shift: Record approved changes to premise, character bios, world rules, or major plot turns promptly. Standalone work asks the author to confirm the update; workflow work returns the proposed change to
write-fiction,revise-fiction,revise-draft,revise-increment, orrevise-requestfor approval and writing to the applicable baseline or story state. - Surface unresolved questions: At the start of new work, list questions left open from prior sessions and ask the user which to address before continuing.
When to Use
This skill is for:
- ✅ Long-form fiction (novels, novellas, short story collections)
- ✅ Multi-chapter manuscripts requiring character/plot consistency
- ✅ Fiction projects needing developmental or line editing
- ✅ Stories with complex worldbuilding or multiple POV characters
When NOT to Use
This skill is NOT for:
- ❌ Flash fiction or single scenes (English guideline only: < 2,000 words) - too lightweight for the workflow
- ❌ Poetry or experimental prose - needs different editorial approach
- ❌ Screenplays or stage plays - different format conventions
- ❌ Technical writing, documentation, or academic papers
- ❌ Business writing or marketing copy
For Chinese work, use an author-approved character count or another author-approved length unit to decide whether the workflow fits. wc -m includes newlines, punctuation, and any Markdown markup, so treat its output as raw inventory metadata unless the author approves that exact counting unit for workflow-fit. Do not derive a Chinese threshold through a fixed English-word conversion.
For narrative nonfiction (memoir, self-help with story elements), use the narrative-nonfiction skill instead.
Editorial Personas
Personas are primarily for Stage 2 (Chapter Development) and Stage 3 (Reader Testing); Stage 1 is freeform Story Bible building that does not require persona invocation. Switch between these roles by requesting a specific lens:
| Role | Invocation | Focus |
|---|---|---|
| Developmental Editor | "As developmental editor..." | Plot, pacing, structure, stakes, theme |
| Line Editor | "As line editor..." | Prose rhythm, word choice, "show don't tell" |
| Character Consultant | "As character consultant..." | Voice consistency, motivation, arc, relationships |
| Continuity Tracker | "As continuity tracker..." | Timeline, world facts, internal consistency |
| Brainstorm Partner | "Brainstorm mode..." | "What if" exploration, problem-solving, unsticking |
Load only the reference file matching the currently invoked persona. Do not preload all references at session start—it wastes context budget. If switching personas mid-session, load the new reference file and treat the prior one as out-of-scope unless the work explicitly bridges both.
See references/ for detailed guidance on each role.
Stage 1: Story Bible Building
Goal: Establish shared story foundation before drafting or editing.
Initial Questions
- Genre and target reader?
- Core premise/logline?
- Protagonist: who they are, what they want?
- Central conflict?
- Reader's intended emotional journey?
- How much written vs. planned?
Story Bible Components
Plot: Premise, three-act structure/beat sheet, major turns, ending (even if rough)
Characters: Protagonist (want/need/wound/arc), antagonist (motivation/threat), supporting cast (function/relationships), POV voice notes
World: Setting (time/place/rules), tech/magic systems, social structures, sensory palette
Theme: Central question, moral argument, recurring motifs
If a Story Bible document exists, review it. If not, offer to create one using assets/story-bible-template.md.
Example Story Bible entry (character):
ALEX CHEN - Protagonist
Want: Expose the conspiracy and clear her name
Need: Learn to trust her instincts over institutional authority
Wound: Mentor betrayed her at previous agency, causing career setback
Arc: Lone wolf → realizes she needs allies → builds trust with team
Voice notes: Analytical, dry humor when stressed, avoids emotional language
Key relationship: Tension with Handler (wants to trust, can't fully)
Exit condition: Confident grasp of story fundamentals. Can discuss character motivations, predict plot implications, and identify thematic threads without asking basic questions.
Stage 2: Chapter Development
Goal: Draft or refine chapters through brainstorm → curate → draft → refine cycles.
Drafting new? → Creation workflow | Editing existing? → Editing workflow
Creation Workflow
-
Scene Planning
- What must happen (plot)? Whose POV?
- Chapter's emotional arc?
- What reader learns/feels by end?
-
Brainstorm Beats (5-15 options): Opening hooks, key moments, dialogue, sensory details, closing
Example (thriller scene): Same car outside coffee shop three days running | Phone buzzing at 3am with blocked caller | Surveillance photo under door | Colleague mentions detail only surveillance would know | Camera lens reflection in window | Dead drop cleaned out | Safe house key doesn't fit | Contact misses first check-in
Then curate: "Which create immediate tension? Combine any?"
-
Curate: Ask which to keep, combine, or discard. Reasons help calibrate.
-
Draft: Write chapter. Use targeted text edits for revisions, never reprint.
-
Refine: Iterate on feedback. After 3 passes with minimal changes, ask: "What could be cut?"
Editing Workflow
-
Read and Diagnose: What chapter tries to do, where it succeeds, where it loses energy/clarity
-
Invoke Persona: Structure/pacing → Developmental | Prose → Line | Voice → Character | Facts → Continuity
-
Check context fit: Load
references/context-fit.mdand complete it before presenting candidates for every concrete edit, including scene rewrites and cross-persona changes. -
Propose Changes: Specific, surgical edits with brief "why" and the checked local context
-
Implement: Use targeted text edits. Link to file after changes.
-
Iterate: Until chapter achieves purpose
Role-Specific Guidance
When a specific editorial persona is invoked, load the corresponding reference file:
- Developmental editing →
references/developmental-editing.md - Line editing in English or other non-Chinese work →
references/line-editing.md - Chinese line editing (
language=zh*or an actually Chinese manuscript) →references/chinese-line-editing.md - Character work →
references/character-work.md - Continuity →
references/continuity-tracking.md - Brainstorming →
references/brainstorming.md - Thriller-specific craft →
references/thriller-craft.md - Sci-fi worldbuilding →
references/scifi-worldbuilding.md
For Chinese work, follow the actual manuscript if its language differs from the label. The Chinese Line Editor must load references/chinese-line-editing.md instead of references/line-editing.md; do not transfer the English said, cut-that, I/The opening, or -ly rules to Chinese. Chinese developmental editing, character work, and continuity tracking keep their existing references and workflow.
Stage 3: Reader Testing
Goal: Verify manuscript works without author context.
Using a fresh reviewer without Story Bible context:
Use a fresh sub-agent when the host provides one. Otherwise, write a handoff containing only the manuscript path, target reader, and review questions, then ask the author to open a new session. The reviewer receives no Story Bible, prior analysis, or revision discussion.
- Comprehension: Can they summarize plot, understand motivations, identify stakes?
- Engagement: Where did they lose interest, have questions, feel confused?
- Emotional: Did key moments land? Ending satisfying? Theme clear?
Common issues: Unclear motivation | Pacing lags | Unearned moments | Confusion
If struggles: Identify gap → Return to Stage 2 → Re-test
Exit condition: Reader understands and engages without author explanations.
Self-Check: Is This Working?
Use these checkpoints to verify you're following the workflow correctly.
After Story Bible building:
- Can you describe the protagonist's want vs. need without re-reading notes?
- Can you predict how the antagonist would react to a new scenario?
- Do you understand the thematic question the book explores?
- Could you summarize the three-act structure in 2-3 sentences?
After invoking a persona:
- Did you explicitly say "As [persona name]..." in your request?
- Is the feedback focused on that persona's domain (developmental = structure, line = prose)?
- Did you avoid mixing feedback from multiple personas in one pass?
After making edits:
- Did you complete
references/context-fit.mdbefore presenting the proposal? - Did you use targeted text edits for surgical changes, not reprinting entire sections?
- Can you articulate what changed and why it's better?
- Is the change consistent with the Story Bible (character voice, plot logic, world rules)?
After brainstorming:
- Did you generate 5+ options before selecting one?
- Did you curate collaboratively rather than taking the first suggestion?
- Can you explain why the selected option is stronger than alternatives?
Before claiming "done":
- Has a fresh reviewer (without Story Bible context) read the manuscript?
- Did the fresh reader understand plot, character motivations, and stakes?
- Were any gaps or confusion points identified and addressed?
If you answered "no" to any checkpoint, return to that stage before proceeding.
Stopping Points
Each persona and stage has a defined end. Stop at it. Do not auto-advance to the next persona, do not silently expand scope, do not start drafting when you were asked to diagnose.
| Tool / Stage | Stop when... | Then |
|---|---|---|
| Story Bible Building (Stage 1) | The Story Bible Self-Check passes | Hand back to author. Do not auto-advance into drafting. |
| Developmental Editor | One full structural pass on the requested scope is complete and issues list is delivered | Wait for author to apply edits. Do not loop into rewriting unless explicitly asked. |
| Line Editor | One chapter (or named scope) is line-edited; after 3 passes with minimal changes, ask "what could be cut?" once | If no further direction, stop. Line editing has a natural ceiling. |
| Character Consultant | One consult on the named character/scene is complete | Do not reflexively check other characters or scenes. Wait for the next invocation. |
| Continuity Tracker | Audit produces flag list | Stop. Do not fix automatically. Author decides which flags are real and how to resolve. |
| Brainstorm Partner | Curate step produces a chosen direction (or 5+ options reviewed and declined) | Hand back. Do not draft into a chapter unless explicitly asked. |
| Reader Testing (Stage 3) | The fresh reviewer's report is delivered by a sub-agent or a separate-session handoff | Stop. Author decides whether to return to Stage 2 and on which findings. |
| Session | Stopping point reached or context window is filling | Record approved Story Bible changes, write a sessions/ note, and stop. |
If the author explicitly asks you to continue past a stopping point, fine—but name what's about to happen ("I'll now move from continuity audit to applying fixes") so the scope shift is visible.
Common Mistakes
| Mistake | Why It Happens | Fix |
|---|---|---|
| Skipping Story Bible | "I know my story well enough" | The Story Bible preserves shared context across sessions. Without it, feedback will miss key story elements. Build it. |
| Generic feedback without persona | Rushing, forgetting to invoke specific role | Explicitly say "As developmental editor..." or "As line editor..." in your prompt. Different lenses catch different issues. |
| Reprinting entire chapters | Habit from other editing contexts | Use targeted text edits for surgical edits only. Reprinting burns context and makes changes hard to track. Link to file after edits. |
| Jumping to line edits before structure | Wanting to "fix" prose immediately | If plot/pacing/character issues exist, line edits are wasted effort. Always developmental pass first. See example below. |
| Skipping Reader Testing | "I've read it so many times already" | You have author context. Reader Testing uses a fresh reviewer without the Story Bible to catch gaps readers will hit. |
| Too many personas at once | Trying to fix everything in one pass | Invoke one persona per pass. Developmental → Character → Line → Continuity. Focused feedback is actionable feedback. |
| Brainstorming without curation | Taking first idea that sounds good | Generate 5-15 options, then curate. First idea is rarely best idea. Quantity enables quality. |
Example: Developmental vs. Line Editing
Same passage, different lenses:
Sarah walked into the office. Her boss looked angry. "We need to talk," he said. She sat down nervously.
Line Editor feedback (prose-level):
- "Walked" is weak—try "strode" or "slipped"
- "Looked angry" tells rather than shows—describe furrowed brow, tight jaw
- "Nervously" is an adverb crutch—show the nervousness through action
Developmental Editor feedback (structure/stakes):
- What does Sarah want in this scene? What does her boss want?
- If this is the confrontation, we need setup—what's the conflict history?
- Stakes feel low—why does this conversation matter to the story?
- Pacing: Is this the right chapter for this confrontation, or should tension build longer?
The difference: Line edits polish sentences. Developmental edits ensure the scene earns its place in the story. Always developmental first.
Quick Reference Commands
| Need | Command |
|---|---|
| Start new project | "Let's build a story bible for [project]" |
| Developmental pass | "As developmental editor, analyze [chapter/section]" |
| Line edit | "As line editor, polish [scene/passage]" |
| Character check | "As character consultant, is [character]'s [action] in character?" |
| Continuity audit | "As continuity tracker, check [chapters X-Y] for inconsistencies" |
| Get unstuck | "Brainstorm mode—I need to [solve problem]" |
| Test readability | "Run a fresh read on [chapter/section]" |
Finding Format
Standalone developmental-editing critiques and continuity-tracking flags conform to references/finding-schema.json: audit = the mode name, technique = the named craft technique, severity, location, issue, confidence: "judgment", and optional exemplar. Serialize findings as JSON, using one JSON code block when embedding them in Markdown; YAML does not satisfy this contract. Brainstorming and drafting proposals remain outside the finding contract.
When write-fiction, revise-fiction, revise-draft, revise-increment, or revise-request supplies workflow report context and the resolved <setup-skill-root>, load <setup-skill-root>/assets/references/workflow-reporting.md and use finding v2 for diagnostics. Return reports and proposed baseline updates without advancing pending state. After a revision orchestrator obtains edit approval, return the affected file list and wait for its backup confirmation before editing; it records backups through author-project.
When write-fiction invokes Chapter Development, use the supplied approved plan and context package. Write only beneath the registered <paths.creation>/staging/<batch-id>/; never write the intended manuscript target, update current story state, or commit the batch. Return each staged path, SHA-256 hash, and intended target path with proposed story-state differences. The caller owns chapter review, acceptance, publication, and scope updates. Standalone behavior remains unchanged: use the normal creation and editing workflows above and leave shared project state untouched.
Files
references/context-fit.md- Proposal-time continuity, history, repetition, voice, intent, and replacement-read checksreferences/developmental-editing.md- Plot, structure, pacing analysis (see also thestory-structureskill for landmark-beat percentages and signpost diagnostics)references/line-editing.md- English and other non-Chinese prose-level refinementreferences/chinese-line-editing.md- Chinese prose-level refinement with contextual language rulesreferences/character-work.md- Voice, motivation, arc tracking (see also thecharacter-archetypesskill for role/personality archetype analysis)references/continuity-tracking.md- Timeline and fact consistencyreferences/brainstorming.md- Idea generation techniquesreferences/thriller-craft.md- Genre-specific guidance for suspensereferences/scifi-worldbuilding.md- Technical accuracy, speculation rulesassets/story-bible-template.md- Blank story bible structureassets/scene-worksheet.md- Scene-level analysis template
How can the creator link this skill?
Add the canonical catalog link to the repository README so users can inspect current installs and available audits. The publishing guide covers the complete discovery path.
<a href="https://skillzs.dev/skills/shengchen1998/author-toolkit/fiction-workshop">View fiction-workshop on skillZs</a>