novel_disassembly_agent_mini

[OMX] Run an anti-slop cleanup/refactor/deslop workflow

[OMX] Ask a local external advisor CLI (Claude or Gemini) and capture a reusable artifact

[OMX] Durable professor-critic research workflow over Codex goal mode without reviving deprecated omx autoresearch

[OMX] Socratic deep interview with mathematical ambiguity gating before execution

[OMX] Setup and configure oh-my-codex using current CLI behavior

[OMX] Run an evaluator-gated performance optimization workflow over Codex goal mode with durable OMX artifacts and safe goal handoffs.

[OMX] Strategic planning with optional interview workflow

[OMX] Self-referential loop until task completion with architect verification

[OMX] Alias for $plan --consensus

[OMX] N coordinated agents on shared task list using tmux-based orchestration

[OMX] Create and execute durable repo-native multi-goal plans over Codex goal mode artifacts.

[OMX] Visual Ralph orchestration for frontend UI from generated references, static references, or live URL targets, using $ralph with built-in visual verdict and pixel-diff evidence until the implementation matches and leaves a reproducible design system.

[OMX] Persistent markdown project wiki stored under repository omx_wiki with keyword search and lifecycle capture

Analyze and improve chapter dialogue design, including speaker differentiation, information efficiency, conflict tone, and reader clarity.

Build a lightweight research/setting pack for chapter continuation, covering world rules, historical analogues, audience expectations, and topic-specific caution points.

Simulate a core web-novel reader review for a chapter draft, focusing on clarity, engagement, expectation, and confusion points.

Inspect chapter rhythm, scene density, tension progression, and ending-hook strength, then return structured issues and repair directions.

Inspect an imitation draft before formal gating and return likely failures, blocking issues, and recommended revision actions.

Gather branch context, rule constraints, relationship state, unresolved threads, and carry-over memory into a machine-readable imitation constraint pack before draft generation.

Distill author-facing lessons from a validated chapter, focusing on hook design, conflict progression, reveal order, scene efficiency, and transferable writing craft.

Review chapter JSON for unsupported inferences, weak evidence coverage, and possible fabricated plot claims.

Generate the analytical layer for a chapter from the factual layer, including summaries, pacing, themes, emotional curve, and continuity notes.

Extract factual layer JSON for a chapter, including characters, events, relations, conflicts, and foreshadowing without inventing unsupported plot facts.

Normalize one chapter into model-friendly structural units such as cleaned text, paragraph blocks, dialogue candidates, scene candidates, and explicit task constraints before downstream extraction.

Bind each extracted fact or analysis claim to explicit source evidence spans, offsets, and confidence so unsupported items can be downgraded or removed.

Convert validated chapter JSON into graph-friendly entity, relation, conflict, and foreshadowing records.

Render a validated chapter JSON payload into deterministic Markdown using local scripts.

Materialize retrieval-friendly chapter documents, chunks, keyword fields, and BM25 input text from validated chapter JSON.