← Search

Story long analyze: deep structural breakdown of long Chinese web novels
Writing 🧪 Install ran in our test sandbox · Nov 18, 2026 📌 Pinned version: commit 7408014
Install test only, on the date shown. This does not guarantee the code is safe.
⚙ Needs: the novel text file to analyze (or an existing…

Story long analyze: deep structural breakdown of long Chinese web novels

A 7-stage pipeline (index → golden chapters → batched extraction → plot/rhythm → characters → report → style) that deconstructs long web novels into per-chapter facts and reusable story mechanics

At a glance
What
A 7-stage pipeline (index → golden chapters → batched extraction → plot/rhythm → characters → report → style) that deconstructs long web novels into per-chapter facts and reusable story mechanics
Cost
Free
Needs
the novel text file to analyze (or an existing analysis to reuse or enhance); Python 3 for the pipeline scripts (the skill tries python3, then python, then py -3, and uses the first that runs); an agent runtime that can spawn subagents for the batched extraction stages (falls back to solo mode otherwise; its spawn check expects agents bundle version 36, and a mismatch is only flagged, never blocking); expect heavy token use on full-novel runs
Install
Copy the installer prompt below into your Muse — your agent does the rest.

Version:

Quick answer

Story Long Analyze is a free Muse skill that deconstructs long web novels into per-chapter facts and story mechanics, through a seven-stage pipeline. It turns a long web novel into per-chapter facts, plot units, character profiles, pacing analyses and a style profile, starting with a chapter index, then golden chapters, batched extraction, plot and rhythm, characters, report and style.

⌁

Install

Copy the install package below, then paste it into Muse🧪 Install ran in our test sandbox · Nov 18, 2026📌 Pinned version: commit 7408014

Step 0, before installing anything: read this install prompt, summarize in a few lines what this build does and what installing it involves, then ask me to confirm before you proceed. Prerequisites: the novel text file to analyze (or an existing analysis to reuse or enhance); Python 3 for the pipeline scripts (the skill tries python3, then python, then py -3, and uses the first that runs); an agent runtime that can spawn subagents for the batched extraction stages (falls back to solo mode otherwise; its spawn check expects agents bundle version 36, and a mismatch is only flagged, never blocking); expect heavy token use on full-novel runs Install "Story long analyze: deep structural breakdown of long Chinese web novels" for me. It gives my agent @zenstory-ai's 7-stage novel-deconstruction pipeline: Stage 0 mechanical chapter index, Stage 1 "golden three chapters" deep dive, Stage 2 batched chapter extraction via subagents (3 batches per round, resume support), Stage 3 plot, dual timelines and 3D rhythm, Stage 4 characters, settings and relationship graphs, Stage 5 master report, Stage 6 style profile, plus an optional inspiration-library pipeline, hard analysis boundaries (unknown when missing, chapter locators, abstract-mechanics-only migration), author-facing plain-language outputs, and resume support. Its Phase 2 now routes by the existing-assets checker's result: a new analysis, continuing a partial one, using a complete one directly, repairing only the missing later stages, or enhancing an incomplete set. Written in Chinese; targets Chinese web fiction. MIT-licensed. Repository: https://raw.githubusercontent.com/zenstory-ai/oh-story-claudecode/74080143dd1ac9a39a930fafd99a0dbbe61b946a/skills/story-long-analyze/SKILL.md (the skill is the whole skills/story-long-analyze/ folder at that commit: SKILL.md plus its references/ and scripts/ subfolders) Pinned version: install the code at commit 74080143dd1ac9a39a930fafd99a0dbbe61b946a exactly (Skill Harbor tested this version). Do not substitute a newer version. If this commit is unreachable, stop and tell me. Updated October 18, 2026: Skill Harbor regenerated this install prompt from the current upstream file. Phase 2 was restructured upstream (75 lines removed, 38 added): instead of a single pipeline table with three cases and full sections for each stage, SKILL.md now routes by the checker's recommended_path (new_analysis, continue_partial, direct_use, repair_stages, enhance_existing), with a second table of stage_repairs codes, and fills only what is missing, in stage order; the detailed Stage 0 to 6 sections are replaced by short per-stage key points that point into the reference files. A new operating conventions section defines the Python placeholder (python3, then python, then py -3), saving a first book's TXT under the analysis folder as 原文/原文.txt and converting non-UTF-8 text (commonly GBK) to UTF-8 first, and suggesting /story-short-analyze when the full text is under 20,000 characters or chapter headings cannot be recognized. The spawn check now expects agents bundle version 36. The fixed delivery paths changed: the golden three chapters are now three separate files, and the plot files are listed under 剧情/, including 剧情/README.md. The installed folder now counts 18 files (SKILL.md, 12 reference files and 5 scripts): references/deconstruction-notes.md, references/semantic-acceptance-fixtures.md and references/style-profile-protocol.md are gone. Correction: the previous prompt's safety check spoke of allowed hosts declared in SKILL.md and credentials only via the secure vault; the new file declares no allowed hosts list and mentions no credentials or vault at all. 1. Fetch the whole skills/story-long-analyze/ folder at the pinned commit, SKILL.md and every file under references/ and scripts/, into a temporary folder, then compare SKILL.md with the fingerprint from Skill Harbor's review of 2026-10-18: - SKILL.md: SHA-256 159f90daf414ff0a9ec298ae5953f4253acf239016e78aa99563ff748eee4a6c If it does not match, it changed since the review: do not install it yet, tell me what you found and let me decide. 2. Safety check: review SKILL.md, its reference files and its scripts for anything suspicious (unexpected network calls, shell commands, credential harvesting); STOP on any red flag and tell me. Tell me in plain words what the skill does when used, which services it connects to, whether it reads or sends any personal information or keys, and anything that seems unusual. Confirm these points from the files: - Agent compatibility and spawning: the skill only checks the current runtime's own agent directory; if the runtime does not support project agents or the files are not found, it degrades to solo/direct and reports "Fallback: project custom agents unavailable -> solo" (ZCode 3.3.4 provides no project custom agents, so it degrades directly). Its spawn version notice reads agents_version from the project root .story-deployed file and expects 36: a mismatch never blocks spawning, it is only reported with advice to rerun /story-setup and open a fresh session (and, above 36, to update oh-story-claudecode first rather than downgrade with a local older setup). - Operating conventions: {PYTHON} means trying python3, python and py -3 in order and using the first that runs; on a first analysis the author's TXT is saved as 原文/原文.txt inside the analysis folder 拆文库/{book title}/, and if its encoding is not UTF-8 (commonly GBK) it is converted to UTF-8 first, so subagents reading by line number do not get garbled text. A full text under 20,000 characters, or chapter headings that cannot be recognized, is probably a short story: suggest /story-short-analyze, unless the author insists or the run was called from an import. - Phase 1: with no book title or source text, ask for the title, platform and source path (a complete existing analysis can be used directly without demanding the source); for an existing folder, run the read-only checker scripts/inspect_existing_assets.py with --compact first. A wrong path must stop the run. Existing chapter summaries are never overwritten, and schema_version is only reported, never used as a gate or rewritten on reuse. - Phase 2 routing: follow the checker's recommended_path: new_analysis (Stage 0 index, Stage 1, stop and ask, then Stages 2 to 6), continue_partial (build the index if missing, fill only the missing Stage 1 products, then plan --intent continue for the missing chapters), direct_use (use a complete existing analysis directly, no index and no reading the source; plan --intent enhance only when the author explicitly asks for enhancement), repair_stages (per-chapter results are complete, fill only the later products listed in the stage_repairs table), enhance_existing (build the index first if the total chapter count is unknown; plan --intent continue fills missing summaries from the old results without reading the source). For stage_repairs codes stage3, stage4, stage5 and stage6, if the products are all present and valid, only mark the stage, do not rewrite them. Stages 3 to 5 use existing results and batch caches only and never reread the full text. - Analysis boundaries and delivery: conclusions come only from readable source text and existing materials, missing facts are written as unknown or not stated in the text, hard facts carry a chapter, a source_locator or a 5 to 15 character locator, inferences are marked with an evidence grade, and only abstract mechanics are migrated. The fixed deliverables include 拆文报告.md (the reading entry), three separate golden-chapter files (章节/第1章_深度拆解.md, 章节/第2章_深度拆解.md, 章节/第3章_深度拆解.md), the plot files under 剧情/ (including 剧情/README.md, 剧情/节奏.md and 剧情/情绪模块.md), chapter_index.csv, _progress.md and _analysis_cache/. Stage 1 stops to ask unless the author asked for a full run, several books at once, or an import auto-continues; Stage 6 reads only 4 to 6 targeted source passages and fails explicitly if no verifiable original sentences can be obtained. The optional inspiration library loads only when the author asks for it and writes only into 灵感库/, without changing the analysis outputs. Nothing is sent anywhere during installation itself. 3. Install it as a skill: copy SKILL.md and its helper files into the agent's skills directory, in a folder named "story-long-analyze", keeping the references/ and scripts/ subfolders as they are. 4. Verify with no network calls: frontmatter valid, files in place. 5. Report what was installed, where, and what I still need to do myself (e.g. point the agent at your novel text file or existing analysis directory, then ask it to run the pipeline stages; the skill is written in Chinese, so work in Chinese when invoking it). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. If anything looks off, stop and ask me.

🔒 Security, provenance and what this touches
🧭 What this touches
Based on the install prompt we scanned, this skill does not run commands, make third-party network calls, install packages, write files, or ask for API keys.
Install prompt analysis · 2026-10-13
🔒 Security and provenance
📌 Pinned version: commit 7408014 · 2026-10-18
🧪 Install ran in our test sandbox · Nov 18, 2026
👁 Observed during the install test · Nov 18, 2026 : registry.npmjs.org
🔎 Last checked · Nov 19, 2026 : no change detected in the files linked by this install prompt
Skill Harbor never reviews the code of a listed skill. These are dated facts about the install prompt and our install test, not a safety certification. Review a skill yourself before you install it. How we verify
Wanna be extra careful?

The install prompt below already includes the vetting steps: your agent follows the community checklist before installing anything with executable code. Want more?

How to check a skill before installing →

✓

Verification certificate

Dated facts about what was checked for this listing, and how. This is not a safety certification.

Pinned version
7408014 · 2026-10-18
Sandbox install test
Install ran in our test sandbox · Nov 18, 2026 · Observed during the install test: registry.npmjs.org
Creator identity
Not verified
Fingerprint watch
Last checked · Nov 19, 2026 : no change detected in the files linked by this install prompt
Install prompt analysis
2026-10-13 : No commands, network calls, packages, file writes or API keys detected.

Skill Harbor never reviews the code of a listed skill. These are dated facts about the install prompt and our install test, not a safety certification. Review a skill yourself before you install it. How we verify

?

Questions

How do I install a skill?

Every product page includes a copy-paste install prompt. Paste it into your Muse and it sets the skill up for you — no manual configuration.

Where does my money go?

Straight to the seller. Skill Harbor never processes payments: checkout happens on the seller’s own page, usually Stripe.

What does the ✓ next to a creator’s name mean?

It means we confirmed the identity of the person behind the listing. It says nothing about the code itself — always check a skill before installing it.