20youssef10/novelforge

Detect overused narrative, dialogue, imagery, character, and plot conventions and propose context-fitting alternatives, while preserving deliberate genre use. Use when prose or structure feels predictable, or when revising familiar beats.

Maintain a reusable author preference profile covering writing goals, dislikes, workflow, planning granularity, autonomy boundaries, language, and feedback preferences across novels. Use to customise how NovelForge works for this author.

Create and audit coherent action and combat sequences through spatial state, visibility, terrain, abilities, resources, costs, objectives, injuries, knowledge, and adaptation. Use for any fight, chase, or physical confrontation.

Classify facts and changes into canon states, run dependency and impact analysis, and gate approved retcons. Use for any change to established canon, any promotion of a proposal to canon, and any decision that alters what has already been written.

Verify that major plot turns arise through coherent prior conditions, triggers, choices, actions, mechanisms, and consequences, distinguishing necessity from coincidence and coincidence from contrivance. Use when a turn feels unearned or arbitrary.

Engineer and audit earned character arcs through governing beliefs, pressure, dilemmas, choices, consequences, reversals, realisations, and end-state transformation. Use when planning or checking whether a character's change is earned.

Design characters as autonomous agents rather than plot tools, tracking identity, psychology, behaviour, relationships, voice, and arc. Use when creating, deepening, auditing, or maintaining any significant character.

Simulate character decisions under goals, beliefs, fears, knowledge limits, relationships, incentives, constraints, and current pressure, in order to test agency and behavioural consistency. Use as a plausibility test for any significant character choice.

Propagate meaningful immediate, delayed, social, material, relational, strategic, political, and thematic consequences through the narrative graph and timeline. Use after any significant change to check that the story acknowledges what it caused.

Enforce a token budget on context retrieval by measuring what a candidate Context Packet would load, ranking sections by decision relevance, and trimming or deferring low-value content. Use before any multi-file task, when a packet feels bloated, or when retrieval has overrun.

Retrieve minimal, targeted context from canon layers, narrative graph neighbourhoods, open obligations, reader and character knowledge, style DNA, author preferences, and branch state before any specialist work. Use for every multi-file task.

Check and maintain relevant continuity across characters, chronology, locations, objects, knowledge, relationships, world rules, abilities, injuries, terminology, and events, without over-tracking detail that cannot affect the story. Use before and after writing whenever consistency matters.

Deliver rigorous, evidence-based, independent critique across structure, causality, agency, continuity, information control, style, originality, clichés, theme, pacing, tension, power, factions, and payoff debt. Use when the author asks for a review or an honest assessment of existing work.

Compose and run a cross-engine audit that selects the right subset of NovelForge engines for a given scope, sequences them in dependency order, and consolidates findings without duplicating work. Use for project-wide reviews, arc audits, pre-publication checks, and diagnosing a novel that fails several dimensions at once.

Write and audit dialogue that maintains distinct character voices, respects knowledge limits, uses subtext, and gives every exchange a conversational function. Use for any conversation, interrogation, negotiation, or spoken conflict.

Model institutions and factions through goals, resources, information, internal blocs, alliances, rivalries, constraints, and strategic responses, allowing partial information and internal disagreement. Use when designing institutions or testing how groups respond to events.

Create, track, audit, and resolve layered narrative setups with stable IDs, visibility and strength assessment, spoiler-risk control, and explicit payoff status. Use for planting clues, auditing payoff debt, and checking reveal integrity.

Detect generic, interchangeable, AI-like prose — abstraction, stock emotional language, empty intensity, formulaic transitions, repeated cadence, low-specificity narration, and exposition substituting for scene — while preserving intentional stylistic repetition. Use when revising or when prose feels interchangeable.

Translate and localise novel content while preserving character voice, tone, humour, cultural context, canonical terminology, invented world terms, and emotional effect. Use for any translation, adaptation, or market-transfer work.

Ingest an existing draft into a NovelForge project by inventorying chapters, detecting POV and tense, extracting characters and relationships, reconstructing the timeline and plot structure, inferring Style DNA, and classifying discovered canon. Use when adopting a manuscript already written, when a project has no Novel Bible, or when retrofitting structure onto finished prose.

Maintain long-term writing-agent memory across sessions using scoped durable notes, retrieval policies, expiry rules, and strict canon boundaries. Use when persisting state beyond a single task, resuming prior work, or preventing stale memory from contaminating a project.

Design and maintain mysteries through truth, clue chains, evidence, hypotheses, suspects, secrets, red herrings, reveal fairness, and post-reveal consequences, keeping author, character, and reader knowledge separate. Use for mystery, investigation, thriller, and psychological suspense.

Maintain the durable graph of entities and typed relationships — characters, events, locations, factions, mysteries, clues, powers, promises, consequences, themes, and motifs — for dependency retrieval and canon-impact analysis. Use when tracing how a change propagates.

Define, maintain, and apply a novel's Style DNA — prose signatures, rhythm, diction, imagery, dialogue balance, description density, viewpoint habits, and stylistic invariants — and compare candidate passages against it. Use for style definition, style-sensitive revision, and voice preservation.

Analyse and preserve narrator identity, narrative distance, focalisation, reliability, and consistency across POV structures, detecting head-hopping, omniscient leakage, and impossible knowledge. Use for any POV design, revision, or audit.

Orchestrate NovelForge across all specialist engines, targeted context, canon governance, style intelligence, versioning, research provenance, and autonomous long-term maintenance. Use for multi-step novel work, project-wide audits, branching, or when the author asks the system to act autonomously.

Plan story architecture at the granularity the project needs, register promises and dependencies, and enforce plan-first drafting. Use for saga, arc, volume, chapter, and scene planning, and before any substantial drafting begins.

Guide a first-time user through starting a novel with NovelForge, from installation check through scaffolding, the first plan, and the first drafted scene. Use when someone reports having just installed NovelForge, asks how to begin, or seems unsure what to do next.

Audit narrative tempo through scene function, event density, decision frequency, information movement, escalation, consequence visibility, and breathing space, rather than word counts. Use when a draft feels rushed, stalled, or flat.

Maintain an indexed ledger of unresolved narrative promises, setups, mysteries, relationship obligations, consequences, and world or power setups awaiting resolution, tracking origin, window, priority, and disposition. Use to prevent forgotten payoffs and overloaded resolutions.

Engineer causally earned plot turns through cause, agency, escalation, reversals, stakes, and consequences. Use when building or auditing any major event, reversal, or structural sequence.

Stress-test power systems for rule loopholes, resource loops, rank bypasses, counter gaps, stacking exploits, contradictions, and progression inconsistencies. Use after defining or changing any power system, and before relying on a system constraint to create tension.

Define and preserve power-system rules, resources, costs, limitations, ranks, progression, techniques, counters, and exceptions, and distinguish raw power from skill, strategy, and circumstance. Use for magic, cultivation, murim, supernatural, and technological capability design and auditing.

Create or extend a NovelForge project by initialising the Novel Bible, copying only the domain templates the novel needs, and recording project configuration such as POV, tense, and approval strictness. Use when starting a new novel, adding a domain to an existing project, or running a novel create command.

Read and update project-level configuration for POV model, tense, approval strictness, autonomy level, and engine selection, so these are project settings rather than facts buried in a plan. Use when configuring a project, changing approval strictness, or when the orchestrator needs to resolve what may proceed without asking.

Run the full NovelForge completion gate across structure, causality, character, information control, style, originality, theme, research, power, faction, continuity, and version dimensions, then report PASS, PASS_WITH_ACCEPTED_ISSUES, or BLOCKED. Use before declaring any scene, chapter, or arc complete.

Model plausible reader knowledge, inference, expectation, uncertainty, and misunderstanding at each narrative point, in order to test mystery fairness, reveal integrity, foreshadowing subtlety, and accidental spoilers. Use whenever information is revealed, withheld, or manipulated.

Research factual and creative material for fiction, separating verified fact, interpretation, speculation, and creative invention, and translating findings into story-useful constraints and detail. Use when a novel depends on real-world accuracy.

Track research sources, claims, verification status, uncertainty, dates, and usage for factual material used in fiction, keeping traceable records of where externally derived facts enter the novel. Use for any claim the story depends on being accurate.

Draft scenes with purpose, conflict, beats, sensory specificity, character agency, and meaningful change. Use for writing or revising any scene once its plan and context are approved.

Manage canon shared across multiple novels in a series, including world bible inheritance, character arcs spanning books, shared glossary and terminology, continuity of chronology, and controlled divergence. Use when a series has more than one novel, when starting a follow-up volume, or when reconciling a book against what earlier books established.

Detect drift away from a novel's established Style DNA across prose, POV, tense, rhythm, diction, atmosphere, dialogue density, imagery, and narrative distance, separating intentional evolution from accidental drift. Use when auditing chapters, revisions, or new material against established voice.

Design and track recurring symbols, motifs, images, objects, colours, places, and sensory patterns with controlled evolution, counter-symbols, and payoff. Use for the figurative layer of a novel and for detecting decorative or over-explained repetition.

Build and audit meaningful tension through stakes, uncertainty, risk, pressure, information gaps, and credible consequences. Use when danger feels hollow, outcomes feel certain, or scenes are nominally dramatic without real stakes.

Track thematic questions, tensions, character embodiments, plot tests, counterexamples, and thematic payoffs, ensuring themes emerge through choices and consequences rather than assertion. Use for thematic planning, integration checks, and ending assessment.

Maintain coherent chronology across dates, durations, ages, sequencing, travel time, flashbacks, and simultaneous events, tracking branch-specific divergences separately. Use when planning, auditing, or changing when anything happens.

Maintain a trope ledger covering role, setup, intended function, reader expectation, subversion status, and resolution, so that genre conventions are used deliberately rather than by default. Use for genre-aware planning and auditing.

Manage story snapshots, branches, experimental alternatives, diffs, comparisons, rollbacks, merges, and canon-safe retcons, keeping branch state isolated until an explicit approved merge. Use before any meaningful edit and for all branching work.

Manage multiple novel projects within one workspace using strict project boundaries, registry metadata, active-project selection, and cross-novel isolation. Use when more than one novel is present, when switching projects, or before any cross-project reference.

Build coherent worlds from geography, history, culture, politics, economics, faith, technology, and social rules, connecting durable facts to the narrative graph and consequences. Use when designing, expanding, or auditing a setting.