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 skills add ttttstc/ni-skill --skill think-like-architectgit clone --depth 1 https://github.com/ttttstc/ni-skillWrote 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/ttttstc/ni-skill/think-like-architect)<a href="https://agentmods.dev/skills/ttttstc/ni-skill/think-like-architect"><img src="https://agentmods.dev/badge/skills/ttttstc/ni-skill/think-like-architect/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/ttttstc/ni-skill/think-like-architect"><img src="https://agentmods.dev/badge/skills/ttttstc/ni-skill/think-like-architect.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00158 | $0.02398 |
| Opus 5.5 | $0.00063 | $0.00959 |
| Sonnet 5.5 | $0.00032 | $0.00480 |
| Haiku 4.5 | $0.00016 | $0.00240 |
Grade A, and why
think-like-architect 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.
How it starts
The opening of the file, as written. The whole thing — 193 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Think Like an Architect
把架构当作当前约束下的高价值决策,不把它当作技术组件清单。
从真实目标、不可变事实、硬约束和成功标准推导方案。先决定高影响、跨边界、难逆转的事项;延迟可逆且等待能获得更多信息的事项。输出可检验、可演进的首层架构假设,不输出最终蓝图。
工作边界
- 只处理会改变系统边界、业务边界、数据所有权、信任边界、质量属性、团队所有权或演进成本的事项。
- 不展开 API 字段、表结构、类、函数、部署脚本和任务拆解,除非某个细节本身构成难逆转的首层决策。
- 不把 PRD 中的技术建议当作事实。把它还原为目标、约束或待验证假设。
- 不依赖
ni-unknown-first,不启动子代理,不自动持久化状态。 - 默认在对话中交付。用户确认导出前,不创建方案文件。
不可跳过的流程
严格按以下状态推进,不得压缩、合并或跳过门禁:
接收上下文
-> 建立事实底座
-> 判断信息充分性
-> 信息不足且可在已授权项目内发现:定向检查项目
-> 信息不足且需超出已授权范围:提交调研提案并停止
-> 信息仍不足:每轮最多提出 3 个高价值问题
-> 提炼架构驱动力
-> 筛选高价值决策
-> 必要时生成结构性不同的选项
-> 形成首层边界、风险和验证路径
-> 输出完整 Architecture First Cut
-> 询问是否导出 Markdown
按阶段加载参考
只加载当前阶段需要的文件,不一次性读取全部参考:
| 当前阶段 | 必须读取 |
|---|---|
| 判断信息是否充分、是否需要项目调研 | references/intake-and-research.md |
| 提炼驱动力、边界和决策 | references/methodology.md |
| 同一高价值决策存在两个以上可行方向 | references/options.md |
| 准备输出完整结论 | references/output-template.md |
第一步:建立事实底座
从用户消息、已提供材料和已授权项目中提取:
- 真实用户、业务价值和待解决问题;
- 成功标准与明确非目标;
- 当前系统、范围和上下游;
- 硬约束:法规、安全、数据、时间、预算、团队和技术环境;
- 已知变化方向与演进期限;
- 事实、假设、Unknown 及其证据来源。
现有系统场景先检查项目文档、代码结构、配置、部署约定、数据与集成边界,再讨论新结构。只检查会影响首层决策的内容,不下钻实现细节。
已授权范围
以下内容视为已授权上下文,可以直接读取:
- 用户粘贴、上传或明确指定的文件、URL 和资料;
- 用户明确要求“基于当前仓库”“分析这个项目”或给出仓库链接/本地项目路径时,该仓库或项目的只读内容;
- 当前对话已经确认的事实和决策。
读取已授权项目不需要重复申请确认。不得把这种授权扩展到其他仓库、外部系统、私有指标、日志、账号或利益相关者资料。
第二步:执行信息充分性门禁
读取 references/intake-and-research.md,逐项判断下列信息是否足以支撑首层决策:
- 目标、用户价值和成功标准;
- 范围、当前状态和明确非目标;
- 硬约束;
- 3–5 个候选业务或质量属性驱动力;
- 数据分类、所有权、安全与信任边界;
- 团队、时间、预算和运维能力;
- 验收标准。
缺失信息只有在可能改变首层决策时才构成阻塞。非阻塞 Unknown 记录后继续,不追求信息完美。
在已授权项目内主动获取
答案可从已授权项目中发现时,主动做最小只读检查:先看架构文档、项目入口、模块边界、配置和部署描述,再按证据缺口缩小范围。不得为了“熟悉项目”无边界扫描。
超出已授权范围的调研确认门禁
需要访问已授权范围之外的资料时,必须先输出调研提案并停止。提案必须说明:
- 哪个首层决策缺少证据;
- 准备获取什么信息、从哪里获取;
- 调研范围与明确不做的内容;
- 预期关闭哪些 Unknown;
- 是否涉及外部访问、敏感数据、额外成本或账号权限;
- 确认问题:
是否授权我按以上范围开始调研?
用户明确确认前,不得访问新来源、启动调研工具或输出正式架构方案。
访谈门禁
项目证据仍不足时,只问会改变高价值决策的问题:
- 每轮最多 3 个问题;
- 先问影响最大且无法从项目发现的问题;
- 说明答案会改变哪类决策;
- 不询问可安全延迟到详细设计的问题。
如果用户拒绝补充或要求立即输出,允许继续,但状态必须为“暂定”,置信度不得标“高”,并显式列出关键假设、阻塞 Unknown 和最便宜的验证动作。
What ships with it
5 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.
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 · 193 lines · 158 tokens per session scan A 3f8a249729c5
think-like-architect is a skill published in the GitHub repository ttttstc/ni-skill (10 stars, last pushed 6d ago), licensed MIT. It adds 158 tokens to every session and 2,398 once invoked, about $0.0006 per session on Opus 5.5. 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 2026-10-30.
think-like-architect compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| think-like-architectttttstc/ni-skill | Skill | 10 | 6d ago | MIT | A | 158 | Codex |
| mem0-oss-to-platformmem0ai/mem0 | Skill | 68k | 2d ago | Apache-2.0 | A | 217 | Claude Code, Codex |
| azure-mgmt-botservice-dotnetmicrosoft/skills | Skill | 3.2k | yesterday | MIT | A | 78 | Claude Code |
| fastapi-router-pymicrosoft/skills | Skill | 3.2k | yesterday | MIT | A | 46 | Claude Code |
| dev-runtimelangwatch/langwatch | Skill | 5.0k | yesterday | Apache-2.0 | A | 172 | Claude Code |
| api-designericrisco/rsc-harness | Skill | 503 | 2d ago | MIT | A | 105 | Claude Code, Codex |
| migrate-segw-to-raparc-mcp/arc-1 | Skill | 252 | yesterday | MIT | A | 106 | Claude Code |