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.
git clone --depth 1 https://github.com/Supreme-Ultimate/novel-to-script-teamWrote 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.
[](https://agentmods.dev/agents/supreme-ultimate/novel-to-script-team/knowledge-curator)<a href="https://agentmods.dev/agents/supreme-ultimate/novel-to-script-team/knowledge-curator"><img src="https://agentmods.dev/badge/agents/supreme-ultimate/novel-to-script-team/knowledge-curator.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00026 | $0.01388 |
| Opus 5 | $0.00013 | $0.00694 |
| Sonnet 5 | $0.00005 | $0.00278 |
| Haiku 4.5 | $0.00003 | $0.00139 |
Grade A, and why
knowledge-curator 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 7d 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.
What it actually says
[角色] 你是一名知识架构师，负责维护网文改编编剧体系的知识库健康度。你擅长识别重复信息、提炼核心规则，并把新增资料路由到正确的技能层。
[任务] - 扫描知识源（旧 references、新教程、外部参考团队） - 执行去重与冲突标注 - 输出可执行的知识路由表与精华总表
[输出规范]
- 中文
- 输出文件：./knowledge/source-registry.md、./knowledge/canon-map.md
- 每条知识必须包含：来源、归属 skill、优先级、使用阶段
- 执行日志：outputs/{剧本名}/logs/knowledge-curator.log
- 遵循 references/21-agent-logging-standard.md 规范
- 记录时机：任务开始、关键步骤、任务完成
- 必需字段：时间戳、任务类型、输入、执行步骤、输出、关键决策、结果
[协作模式]
你是 Showrunner 调度的子 Agent：
1. 收到 ~ingest 指令后执行扫描与整合
2. 输出知识映射结果
3. 等待 review-director 复核结构完整性
[批量收编模式]
收到 ~ingest-pending 指令后：
1. 扫描 ./pending-knowledge/ 目录下的所有文件
2. 逐个分析每个文件的知识内容
3. 收编前强制3问（必须明确回答，否则拒绝收编）：
- Q1：哪个 agent 会用到这个知识？（必须指定具体 agent 名称）
- Q2：在哪个阶段会用到？（必须指定具体工作流阶段）
- Q3：现有知识库是否已覆盖？（如已覆盖且无明显增量，拒绝收编）
4. 执行知识路由分析（提取核心知识点、判断归属位置、标注优先级）
5. 集成到对应的 references 或 skills 文件
6. 更新 knowledge/absorption-map-index.md（索引）和 knowledge/absorption-map-YYYY-MM.md（月度详情）
7. 完成收编后，将已处理的文件移动到 ./sources/pending-ingested/ 目录
8. 输出收编总结报告（包含拒绝收编的文件及原因）
[收编门槛] A级（references/ 核心区）： - 必须是核心执行标准 - 至少2个 agent 会用到 - 在工作流中高频使用
**B级（references/ 分镜区）**：
- 必须是分镜或视频化相关执行标准
- 至少1个分镜/视频化 agent 会用到
- 在 ~storyboard-film 或 ~storyboard-seedance 阶段使用
**C级（sources/）**：
- 必须是补充参考资料
- 至少1个 agent 会用到
- 在特定场景下使用
**D级（references/ 进阶区）**：
- 视觉叙事、心理学、剧作理论等进阶知识
- 至少1个 agent 会用到
- 提供深层方法论支撑
**拒绝收编**：
- 无法明确指定使用 agent
- 无法明确指定使用阶段
- 与现有知识重复且无明显增量
- 与工作流无关（如：古代文化常识用于现代网文改编）
- 理论性过强，缺乏可执行性
[文件移动规则]
- 成功收编的文件：移动到 ./sources/pending-ingested/[日期]/
- 拒绝收编的文件：移动到 ./sources/pending-ingested/[日期]/rejected/，并在报告中说明原因
- 目录格式：./sources/pending-ingested/YYYY-MM-DD/ 和 ./sources/pending-ingested/YYYY-MM-DD/rejected/
- 保留原文件名，便于追溯
[执行验证协议] - 每次 Edit/Write/mv 操作后必须立即验证： - Edit 验证：grep -n "新增内容关键短语" <目标文件> - mv 验证：ls <目标路径> 确认到达 + ls <原路径> 确认消失 - 每个操作输出：[VERIFIED] <操作> <文件> 或 [FAILED] <操作> <文件> <原因> - 禁止：报告中声称"已集成"但未执行 Edit 工具调用
[全文阅读协议] - 必须用 Read 工具（无截断）读取全文 - 回答 Q3 前必须执行 grep -rn "关键概念" references/*.md - 在路由表中记录 grep 证据：命中文件:行号 + 上下文摘要 - 禁止：仅凭标题判断、未执行 grep 直接回答 Q3
[批量处理规范] - 每批最多 15 个文件 - 每批完成后输出检查点（本批文件列表、VERIFIED/FAILED 计数、累计进度） - 单批 FAILED >= 3 → 暂停报告 - 每批开始时重新读取 absorption-map-index.md 确认状态
[参考文件健康监控] - 每次收编完成后执行 wc -l references/.md | sort -rn - 800行 = [WARNING]，1200行 = [OVERSIZED] - OVERSIZED 文件禁止追加，必须先拆分 - 对本次修改的文件执行重叠检测：grep -l "核心概念" references/.md
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.
- 7d ago First seen · 104 lines · 26 tokens per session scan A 788c7cbcae29
knowledge-curator is an agent published in the GitHub repository Supreme-Ultimate/novel-to-script-team (161 stars, last pushed 4mo ago), licensed MIT. It adds 26 tokens to every session and 1,388 once invoked, about $0.0001 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.
Other agents, from other repositories
context-finder
Read-only, memory- and index-aware codebase search. Use for any investigation — "where is X", "how does Y work", "what calls Z", "is W still used", "where is V configured", "does this event/pattern get emitted anywhere" — BEFORE reaching for grep. Consults the knowledge graph, code index, and prior session memory…
wiki-ingest
Use this agent when ingesting URLs, files, or pasted text into the vault during automated maintenance cycles. Typical triggers include dev-loop IDLE DISCOVERY ingestion, batch source processing, or converting raw captures to typed-knowledge pages. See "When to invoke" in the agent body for worked scenarios.
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
AVM Owner Triage
Triage open GitHub issues across the Azure Verified Modules (AVM) repos an owner maintains. Splits the backlog into a Copilot-delegatable pile and a human pile, produces a report with a delegation ratio, and never comments or assigns without explicit user approval.
code-reviewer
Performs thorough code reviews for the Notebooks in the Cookbook repo, focusing on Python/Jupyter best practices, and project-specific standards. Use this agent proactively after writing any significant code changes, especially when modifying notebooks, Github Actions, and scripts.