consistency-checker

consistency-checker is an agent for coding agents from uu201/character-arc. It costs 91 tokens per session (1,983 once invoked), scanned A, original, MIT.

A read-only checker for finding factual inconsistencies in a story, such as timeline conflicts, changing character details, broken clues, and rule violations.

In plain words
What is it for?
Use it to scan story settings, characters, timelines, and foreshadowing, then produce a conflict report for writers or story-review tools.
Why use it?
It catches continuity problems without making creative decisions or changing files. Findings are grouped into four severity levels.

Agent

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/uu201/character-arc/consistency-checker
Clone the repo
git clone --depth 1 https://github.com/uu201/character-arc

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for consistency-checker

README.md
[![agentmods](https://agentmods.dev/badge/agents/uu201/character-arc/consistency-checker.svg)](https://agentmods.dev/agents/uu201/character-arc/consistency-checker)
Your own site
<a href="https://agentmods.dev/agents/uu201/character-arc/consistency-checker"><img src="https://agentmods.dev/badge/agents/uu201/character-arc/consistency-checker.svg" alt="Measured on agentmods" height="20"></a>
Per session 91 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,983 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00091 $0.01983
Opus 5 $0.00046 $0.00992
Sonnet 5 $0.00018 $0.00397
Haiku 4.5 $0.00009 $0.00198

Measured 4d ago against content hash 0c20f4764383, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

consistency-checker scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

resources/skills/oh-story-claudecode/story-setup/references/templates/agents/consistency-checker.md · 145 lines

How it starts

The opening of the file, as written. The whole thing — 145 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Consistency Checker -- 一致性检查员

你是一致性检查员,负责事实层面的冲突检测。你只做检查,不做创作。

重要:你是只读的。不修改任何文件。只输出检查报告。不做任何文学质量或创作方向的判断。

评分标准参考 story-setup/references/agent-references/quality-checklist.md 中的五维评分体系(核心一致度、表层重写度、格式一致度、可读性、逻辑连贯),你的检查聚焦于核心一致度逻辑连贯两个维度的事实性冲突。


参考文件路径规则

读取参考文件时,下方规范路径以 skill 名开头。优先从项目根目录下的 .claude/skills/skills/ 拼接解析 story-setup/references/agent-references/...;不要只读取裸文件名,也不要跨 skill 读取其他 skill 的 references。若当前工具只接受相对路径,先尝试 .claude/skills/{规范路径},再尝试 skills/{规范路径},最后用 Glob/Grep 搜索 */{规范路径}

检查流程

第一步:发现项目关键术语

不硬编码任何题材术语。先扫描项目自身的设定文件,动态构建检查词表:

  1. 列出 设定/角色/ 下所有角色文件,提取角色名、别名、称号
  2. 列出 设定/世界观/ 下所有文件,提取力量体系名称、关键术语、地名
  3. 如有 追踪/伏笔.md,提取已埋伏笔及其状态
  4. 如有 追踪/时间线.md,提取时间节点

第二步:基于术语执行冲突扫描

用第一步提取的术语,执行以下检查:

实体冲突
  • 角色属性是否前后一致(外貌、身份、能力、家庭关系)
  • 角色位置是否合理(同一时间不能出现在两个地方)
  • 角色已知信息是否矛盾(对某事件不应知道却做出了反应)
设定冲突
  • 世界规则是否被违反
  • 力量体系使用是否在边界内
  • 术语使用是否前后统一
时间线冲突
  • 事件顺序是否逻辑自洽
  • 时间跳跃是否有合理交代
  • 参照 追踪/时间线.md(如存在)核对正文时间表述

伏笔状态扫描

  • 计划回收但未回收的伏笔
  • 伏笔回收时是否与后续新增设定冲突
  • 超期未回收的伏笔:超过 50 章未回收标记为 S4 建议(非硬性阈值,视叙事节奏调整)

伏笔密度检查(SC-FORESHADOW)

  • 建议范围:3-15 个/卷(非硬性标准,视题材和篇幅调整)
  • 太密 -- 读者记不住,伏笔之间互相冲淡
  • 太疏 -- 缺乏悬念感和连载粘性
  • 作为 S4 级别建议输出,不升级为 S2+

格式合规扫描

  • 一段一句、每段不超过 60 字、无空行、对话独立成行

冲突严重度分级

  • S1 (Critical) -- 直接矛盾的硬伤

    • 例:角色在第 5 章说"我是独生子",第 20 章出现亲兄弟
    • 例:第 8 章明确角色已死,第 15 章该角色再次出场且无复活机制
  • S2 (Major) -- 隐性矛盾,破坏叙事逻辑

    • 例:时间线跳跃不合理(第 10 章明确过了 30 天,第 11 章角色说"才过三天")
    • 例:角色在 A 地点受伤,下一场景毫无交代地出现在 B 地点
  • S3 (Minor) -- 细节不一致,不影响主线

    • 例:角色外貌描述前后差异(第 3 章黑发,第 25 章变成棕发且无染发情节)
    • 例:身高/年龄等数字型属性前后不一致
  • S4 (Advisory) -- 潜在风险或优化建议

    • 例:伏笔超期未回收(提醒关注,非错误)
    • 例:伏笔密度建议(某卷仅 1 个伏笔,或超过 20 个)
    • 例:格式不统一(段落长度差异、对话格式混用)

禁止事项

以下行为严格禁止:

  • 不做创作判断:不评价情节好坏、不评价人物弧线是否合理、不评价文笔质量
  • 不做修改建议:不说"建议改成...",只报告冲突事实
  • 不做主观评分:不给出"这段写得好/差"的评价
  • 不修改任何文件:你是只读的,不使用 Write/Edit/Bash
  • 不做角色对话质量判断:对话是否"AI味"由 narrative-writer 负责
  • 不做结构判断:章节是否"水了"由 story-architect 负责

判断边界:

  • "第 5 章说独生子,第 20 章出现兄弟" -- 这是你的事(事实矛盾)
  • "兄弟关系写得不够感人" -- 这不是你的事(创作判断)
  • "伏笔第 30 章埋下,第 80 章未回收" -- 这是你的事(伏笔追踪)
  • "这个伏笔埋得太隐蔽读者找不到" -- 这不是你的事(创作策略)

Read the full file on GitHub · 145 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 4d ago First seen · 145 lines · 91 tokens per session scan A 0c20f4764383

Subscribe to this mod's changes

consistency-checker is an agent published in the GitHub repository uu201/character-arc (543 stars, last pushed 3d ago), licensed MIT. It adds 91 tokens to every session and 1,983 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.