worldsmith

Apply a change to the worldbuilding project — edits, new ideas, or promotions

Run read-only diagnostics on the worldbuilding project. Do NOT modify any files. Report findings and let the user decide what to act on.

Show worldsmith commands, skills, and typical workflows

Scaffold or adopt a worldbuilding documentation ecosystem

Launch multi-agent editorial review — consistency, craft, voice, and structure auditors in parallel

Specialist agent for character voice, arcs, and relationships. Launched by the writer orchestrator during multi-agent content generation. Develops character documentation with behavioral specificity — voice patterns testable against dialogue, emotional flickers anchored to moments, and relationship maps with bidirectional behavioral signatures. <example> Context: Orchestrator needs voice patterns developed for a protagonist. user: "Develop Maren's voice patterns and speech tics so the scene-writer can write her dialogue" assistant: "I'll launch the character-developer to build Maren's voice spec — rhetorical habits, stress markers, metaphor families, and patterns specific enough that the voice-auditor can verify dialogue against them." </example> <example> Context: Orchestrator needs a character arc mapped with emotional flickers. user: "Map Kael's arc from chapter 3 through chapter 12 with specific emotional beats" assistant: "I'll launch the character-developer to trace Kael's trajectory — anchored emotional flickers at each chapter, the setbacks that make the arc feel earned, and notes for the scene-writer on how his voice shifts as the arc progresses." </example>

Specialist agent for objective consistency verification in fiction manuscripts. Launched by the reviewer orchestrator during multi-agent review. Checks timeline, factual, character state, and spatial consistency against canonical documentation. Does not evaluate prose quality, voice, or structure. <example> Context: Orchestrator needs timeline and factual consistency verification. user: "Check this manuscript against the canonical docs for contradictions" assistant: "I'll launch the consistency-auditor to verify timeline, facts, character state, and spatial consistency." </example> <example> Context: Orchestrator needs to verify manuscript after canonical doc changes. user: "The timeline was updated — check the manuscript for new contradictions" assistant: "I'll launch the consistency-auditor to compare the manuscript against the updated canonical docs." </example>

Specialist agent for prose craft analysis in fiction manuscripts. Launched by the reviewer orchestrator during multi-agent review. Finds show-don't-tell violations, dialogue craft failures, sentence-level weaknesses, and scene mechanic problems. Runs mechanical pattern counts via count_patterns.py and layers analytical judgment on top. Does not evaluate consistency, voice, or structure — only prose quality. <example> Context: Orchestrator needs prose craft analysis during multi-agent review. user: "Audit the manuscript for prose craft issues — cliches, telling, weak dialogue" assistant: "I'll launch the craft-auditor to analyze prose quality: show-don't-tell, dialogue craft, sentence mechanics, and scene structure." </example> <example> Context: Orchestrator needs pattern counts and craft analysis after a major revision pass. user: "Run a full prose audit including pattern counts on the revised chapters" assistant: "I'll launch the craft-auditor to run count_patterns.py and perform analytical prose review on the revised manuscript." </example>

Specialist agent for worldbuilding content generation — history, mythology, cultures, and systems. Launched by the writer orchestrator during multi-agent content generation. Develops canonical documentation with narrative prose quality and consequence chains. <example> Context: Orchestrator needs world history developed for a kingdom. user: "Develop the history of the Northern Kingdom from founding through the civil war" assistant: "I'll launch the lore-writer to build the Northern Kingdom's history — geological constraints, founding myths, political evolution, and the civil war's causes and aftermath." </example> <example> Context: Orchestrator needs a magic system designed with full consequences. user: "Design a resonance-based magic system and derive its societal implications" assistant: "I'll launch the lore-writer to design the resonance magic system — mechanics, costs, power structures, economic implications, and cultural attitudes." </example>

Multi-agent fiction review orchestrator. Acts as editorial director: reads the project, spawns specialist reviewers (parallel), cross-verifies critical findings, and synthesizes a unified editorial report to .worldsmith/reviews/. <example> Context: User wants a thorough autonomous review of their fiction manuscript. user: "Do a thorough review of my manuscript" assistant: "I'll launch the reviewer agent for a comprehensive multi-agent editorial review." </example> <example> Context: User wants specific chapters reviewed. user: "Critique chapters 3-5" assistant: "I'll launch the reviewer agent to run all four specialists against chapters 3 through 5." </example> <example> Context: User wants a full consistency and craft pass before publishing. user: "Review my novel for consistency and craft issues" assistant: "I'll launch the reviewer agent for a full editorial review — consistency, craft, voice, and structure." </example> <example> Context: User wants a comprehensive editorial report. user: "Full editorial review" assistant: "I'll launch the reviewer agent to orchestrate all four specialist auditors and produce a unified report." </example>

Multi-agent fiction revision orchestrator. Acts as editorial implementer: reads review findings from .worldsmith/reviews/, dispatches writer specialists to fix issues, then dispatches reviewer specialists to verify each fix resolved the original finding. Bridges the reviewer (diagnoses) and writer (creates) with a fix-then-verify feedback loop. <example> Context: User has a review report and wants the issues fixed. user: "Fix the issues from the review" assistant: "I'll launch the rewriter agent to implement fixes from the review findings, with verification after each fix." </example> <example> Context: User wants to revise specific chapters based on editorial feedback. user: "Revise chapters 3-5 based on the editorial report" assistant: "I'll launch the rewriter agent to fix the findings for chapters 3 through 5, verifying each fix with the relevant reviewer specialist." </example> <example> Context: User ran a review and wants targeted fixes. user: "Fix just the HIGH consistency issues from the last review" assistant: "I'll launch the rewriter agent to fix the HIGH consistency findings, with verification by the consistency-auditor." </example> <example> Context: User wants the full review-then-fix cycle. user: "Review my manuscript and fix what you find" assistant: "I'll launch the reviewer first for a full editorial review, then the rewriter to implement the fixes with verification." </example>

Specialist agent for prose scene drafting with craft discipline. Launched by the writer orchestrator during multi-agent content generation. Produces draft prose that shows rather than tells, with distinct character voices, concrete sensory detail, and scene structure that creates and releases tension. <example> Context: Orchestrator needs a confrontation scene drafted for chapter 7. user: "Draft the scene where Maren confronts her mother about the treaty" assistant: "I'll launch the scene-writer to draft the confrontation — entering late, writing to Maren's documented voice patterns, and grounding the tension in physical detail." </example> <example> Context: Orchestrator needs a quiet character scene between two chapters. user: "Write the interlude scene where Kael discovers the broken resonance stone" assistant: "I'll launch the scene-writer to draft the discovery scene — sensory detail, Kael's documented speech tics, and a turn that shifts the stakes." </example>

Specialist agent for pacing, tension, scene turns, thematic coherence, and arc trajectory in fiction manuscripts. Launched by the reviewer orchestrator during multi-agent review. Evaluates whether the narrative works at scene and chapter level — the structural problems that make a reader put a book down without knowing exactly why. Does not evaluate consistency, prose craft, or voice — only structure. <example> Context: Orchestrator needs structural analysis during multi-agent review. user: "Audit the manuscript for pacing problems, weak scene turns, and arc trajectory" assistant: "I'll launch the structure-auditor to evaluate pacing, tension, scene turns, thematic coherence, and character arc trajectory." </example> <example> Context: Orchestrator needs structure audit after major plot restructuring. user: "We reordered the chapters and added new scenes — check whether the pacing and arcs still work" assistant: "I'll launch the structure-auditor to verify scene balance, tension distribution, and arc progression after the restructure." </example>

Specialist agent for character voice, dialogue distinctiveness, and POV consistency in fiction manuscripts. Launched by the reviewer orchestrator during multi-agent review. Verifies characters sound like themselves and like distinct people — the hardest thing for AI-assisted fiction to get right. Does not evaluate consistency, prose craft, or structure — only voice. <example> Context: Orchestrator needs character voice verification during multi-agent review. user: "Audit whether each character's dialogue matches their documented voice patterns" assistant: "I'll launch the voice-auditor to verify voice consistency, distinctiveness, and POV discipline against the character docs." </example> <example> Context: Orchestrator needs voice audit after adding new scenes with multiple characters. user: "Check whether the new chapters maintain distinct character voices or if they're collapsing" assistant: "I'll launch the voice-auditor to compare dialogue and internal monologue against character voice specs and flag any convergence." </example>

Multi-agent fiction writing orchestrator. Acts as lead author: reads the project, plans assignments, delegates to specialist writers (lore, scene, character) in parallel, integrates their output, propagates changes through the canonical doc ecosystem, and ensures everything stays consistent. <example> Context: User wants a chapter drafted from their outline. user: "Write chapter 5" assistant: "I'll launch the writer agent to orchestrate multi-agent content generation for chapter 5." </example> <example> Context: User wants a specific scene written. user: "Draft the battle scene at Greymoor" assistant: "I'll launch the writer agent to coordinate lore, scene, and character specialists for the Greymoor battle." </example> <example> Context: User wants worldbuilding content developed. user: "Develop the Ashwalker culture" assistant: "I'll launch the writer agent to build out the Ashwalker culture — history, customs, internal diversity, and consequences." </example> <example> Context: User wants character work done. user: "Flesh out Sera's backstory" assistant: "I'll launch the writer agent to develop Sera's backstory — voice patterns, arc trajectory, and relationship dynamics." </example>

Use when writing fiction, editing fiction prose, reviewing chapters, or drafting scenes. Covers prose craft, dialogue mechanics, scene structure, and common failure modes in AI-assisted fiction. Trigger phrases: "write fiction", "edit chapter", "prose review", "scene draft", "dialogue check", "fiction craft", "edit this scene", "review my prose".

This skill should be used when the user asks about "documentation structure", "doc relationships", "cross-references", "propagation", "canonical workflow", "exploratory workflow", "docs-first", "lore management", "worldbuilding docs", "consistency rules", "canonical hierarchy", "update docs", "change propagation", "editorial workflow", "Silmarillion approach", or is working in a project with a worldbuilding documentation ecosystem (docs/ or lore/ directory).

4 hooks across 3 events