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.
npx agentmods add skills/leobai03/tc/tc-knowledgenpx skills add Leobai03/tc --skill tc-knowledgegit clone --depth 1 https://github.com/Leobai03/tcWrote 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/skills/leobai03/tc/tc-knowledge)<a href="https://agentmods.dev/skills/leobai03/tc/tc-knowledge"><img src="https://agentmods.dev/badge/skills/leobai03/tc/tc-knowledge.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 | $0.00150 | $0.02105 |
| Opus 5 | $0.00075 | $0.01052 |
| Sonnet 5 | $0.00030 | $0.00421 |
| Haiku 4.5 | $0.00015 | $0.00211 |
Grade A, and why
tc-knowledge 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 5d 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.
How it starts
The opening of the file, as written. The whole thing — 157 lines — stays where its author put it; the contents beside it link to each section on GitHub.
TC Knowledge|天策知识库
一句话说明
知识库像一个有管理员的图书馆:核心参考源是看问题的眼镜,原推是原书,知识原子是带出处的小卡片,知识包是按问题装好的教材。不能把书架上所有句子都当成正确答案。
六种模式
1. 查 TC 知识
用户问天策过去说过什么、某个方法来自哪里,或要求结合 TC 知识时,使用本 Skill 目录内的脚本:
python3 scripts/tc_knowledge.py search --query "目标用户 产品化" --scope guidance --limit 5
实际运行时从本 SKILL.md 所在目录解析脚本绝对路径,不假设当前工作目录。搜索范围:
guidance:默认范围,只查核心参考源、知识包和知识原子;atoms:经过筛选、带边界的知识原子;posts:仅用于用户明确要求的历史查询,只证明作者当时公开说过;packs:可以直接用于判断和行动的专项知识包;sources:技术合伙人与天策的两份核心参考框架;all:兼容旧调用,与guidance相同,不包含历史原推。
回答创业方法时先选相关核心参考源,再用知识包和原子落地。不得为了补故事、证明方向或模仿作者而检索原推。
只有用户明确问“天策过去说过什么、某条原推是什么、历史上怎样表达”时,才使用 --scope posts。返回后停留在历史查询,不把作者的职业、项目、收入、失败、违规经历或其他个人经历带回 /tc 生成项目、推荐方向或推断用户能力。
2. 接用户自己的知识库
用户提供飞书、Notion、网盘、代码仓库或本地目录时:
- 先找项目级
AGENTS.md、knowledge/SOURCE_OF_TRUTH.md或SOURCE_OF_TRUTH.md; - 没有导航时,只扫描用户指定范围,先列资料地图;
- 建立一份短导航,说明不同问题去哪里找、更新时间、是否允许公开;
- 具体回答继续读取导航指向的原始文件,不只根据摘要作答;
- 用户没有要求修改时,只读,不自动整理或上传资料。
对用户统一叫“知识库”和“知识库导航”,不要求他理解 RAG、向量库或真源等术语。
3. 把新资料放进 TC
先登记来源,再按五层处理:
L0 原始资料:保存出处和权限,不代表认可
L1 候选原子:有场景、动作、边界和可证伪条件
L2 已验证模式:拿到真实动作与结果,或多次出现
L3 稳定方法:脱敏、通用、合规、跨案例稳定
核心参考源:多条方法背后相对稳定的世界观,只用于形成问题和判断视角
知识对象使用 QST / CON / OPI / CAS / SOL。一次先提炼 5 至 10 条高价值样本;结构和去重规则稳定后才扩大,不用“全量导入”冒充知识工程。
第三方资料没有公开授权时,原文、私有链接、人物、客户、收入和经营细节只留内部;公开版只保留脱敏后的通用方法。被列为“核心参考源”不等于作者每句话都正确,也不提高其中事实的证据等级。
4. 检查知识库
在官方 TC 安装包中运行:
python3 scripts/tc_knowledge.py stats
python3 scripts/tc_knowledge.py validate
在用户知识库中检查:导航路径是否存在、多个版本是否冲突、动态事实是否过期、私密资料是否被错误标成可公开、答案是否能回到原文件。
5. 查或加工 DBS 外部理论库
用户明确要求使用 dontbesilent 的开源推文集、DBS books 或“那 1,544 条精选推文”时,必须先完整读取 dbs-books.md。AI 只用 Markdown,不读 PDF。
第一次使用时,说明上游是 CC BY-NC 4.0;用户确认仅作署名的非商业研究后运行:
python3 scripts/tc_knowledge.py external-sync \
--source dbs-books --accept-license
然后按当前真问题搜索,不要顺序阅读 1,544 条:
python3 scripts/tc_knowledge.py search \
--scope dbs-books --query "流量 变现 产品" --limit 5
已经下载原文件时,用 --source-path 指向 Markdown 文件或其目录。返回结果必须保留日期、原帖链接、主题、标签、第三方观点和许可证边界。
What ships with it
15 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- agents/openai.yaml 324 B
- references/atoms.jsonl 16 KB
- references/concept-dictionary.md 2.4 KB
- references/core-sources/天策_公开实践与创业价值观.md 2.4 KB
- references/core-sources/技术合伙人_商业世界观与AI实操.md 5.4 KB
- references/external-sources/dbs-books.md 4.4 KB
- references/knowledge-packs/action_市场验证与复盘.md 1.2 KB
- references/knowledge-packs/ai_AI商业落地.md 1.7 KB
- references/knowledge-packs/community_组织与合作边界.md 1.1 KB
- references/knowledge-packs/content_个人IP与内容获客.md 1.2 KB
- references/knowledge-packs/diagnosis_真问题与商业断点.md 1.5 KB
- references/knowledge-packs/product_需求产品化与交付.md 1.1 KB
- references/public-posts.jsonl 521 KB
- references/README.md 1.1 KB
- scripts/tc_knowledge.py 32 KB runs code
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.
- 5d ago First seen · 157 lines · 150 tokens per session scan A 0d0f89c54cd3
tc-knowledge is a skill published in the GitHub repository Leobai03/tc (22 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 150 tokens to every session and 2,105 once invoked, about $0.0007 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 skills, from other repositories
knowledge-base-management
Obsidian 知识库全生命周期管理:三层架构、素材入库(ABC分级)、健康检查、GBrain/GraphRAG/LLM Wiki 三件套集成、目录整理.
make-skill
用户调用了 /make-skill。你的任务是通过对话引导用户,把当前会话中解决问题的方式沉淀为一个可复用的 Scream Code Skill,并安装到插件中心。.
agent-memory-mcp
A hybrid memory system that provides persistent, searchable knowledge management for AI agents (Architecture, Patterns, Decisions).
21-day-self-interview
夜间自我访谈引导(21 天)。扮演资深存在主义心理咨询师,每晚提三个有意义的问题,记录回答并在节点回映,帮助用户看清自己。Use when running a nightly self-inquiry / journaling ritual driven by a scheduled task.
memory-log
查看记忆层活动流水(promotion / migration / decay / correction / memory rollback)——从 events.jsonl 重组时间线,并用 memory-index sidecar 补充每条的当前状态。只读,不改记忆。.
obsidian-doc-structure
用 Obsidian 官方 CLI 读取文档 outline、properties、property、tags、aliases、wordcount——比 Read 快且不把全文灌进上下文。CLI 不可用时 fallback 到 Read + Markdown/frontmatter 解析。.