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 instructions/taichuy/1flowbase/agents-mdgit clone --depth 1 https://github.com/taichuy/1flowbaseWhat 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 | $0.01444 | $0.01444 |
| Opus 5 | $0.00722 | $0.00722 |
| Sonnet 5 | $0.00289 | $0.00289 |
| Haiku 4.5 | $0.00144 | $0.00144 |
Grade A, and why
1flowbase AGENTS.md 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 2d 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 — 59 lines — stays where its author put it; the contents beside it link to each section on GitHub.
记忆
命中对应记忆存储规则自动更新对应记忆中
@.memory/AGENTS.md
用户偏好
.memory/user-memory.md 是本地私有文件，不提交；缺失时参考 .memory/user-memory.template.md 初始化。
本项目相关skill在
.agents/skills 是项目 skill 源目录,其他skills只作为本地镜像 如果没有注册，请自行更新到对应约定目录
本项目 skills
1.problem-framing：需求对齐、会导向产品决策的功能 / 交互诊断、方案拍板与计划形态选择；计划只使用 Single Issue 或两层 Issue Tree。
2.frontend-development：前端页面、UI 结构、工作区流程、节点开发、schema UI、交互和视觉结构变更时使用。
3.backend-development：后端 API、状态流转、模块边界、核心业务逻辑、状态写入口和一致性设计变更时使用。
4.test-driven-development：功能、缺陷、重构或行为变化可用自动化测试覆盖时，在实现前使用。
5.qa-evaluation：进入自检、验收、回归、交付或质量评估阶段时使用，输出证据驱动的 QA 结论。
6.github-solution-research：具体工程问题可能已有 GitHub 开源证据或方案时使用；搜索 issue、PR、repo、code，比较候选项目并提炼本地适配与验证计划。
7.mcp-configuration-development：从当前前后端源码构建、修复和审计 MCP Virtual UI 配置，完成 Tool、mapping、Group、Binding、discovery policy 与 mcp_list/get/call 验证；不用于修改 MCP 运行时代码或决定产品语义。
质量控制
1.进入自检、验收、回归或交付阶段时，使用skill qa-evaluation；
2.前端实现规则: web/AGENTS.md
3.后端实现规则: api/AGENTS.md
4.warning 与 coverage 产物统一落到 tmp/test-governance/。
5.涉及功能、缺陷、重构或行为变化的开发，先使用项目 skill test-driven-development；若不适用，交付说明需写明原因与替代验证。
6.后端是唯一数据来源，前端不应该作代码处理输出兼容，应该是后端提供职责单一的接口
7.前后端接口字段名必须与后端 DTO / 领域语义保持一致；UI 展示名可以本地化，但不得为展示文案另起接口字段别名。确需兼容旧字段时，必须在兼容代码最近行标记 @field-contract-compat source=... alias=... remove_by=yyyy-mm-dd，带废弃计划和测试，并让 repo-hygiene 以 warning 暴露。
8.前端多语言资源必须让 i18n-hygiene 暴露无静态引用 key；unused-i18n-key 是 warning，保留时必须说明动态 key、路由配置或外部渲染入口原因。
9.受保护页面和认证 API 的本地运行态取证使用 page-debug / api-debug；临时脚本不得直接调用 /api/public/auth/sign-in。确需自定义 Playwright 流程时，复用 scripts/node/page-debug/auth.js 的临时 session owner，并在 finally 回收。
开发流程控制
1.需求类请求，以及会导向产品决策的功能差异、缺失能力、不可编辑状态或入口归属诊断，默认先使用 problem-framing；方向确认前不修改产品代码。
2.普通任务使用 Single Issue；跨上下文、多 agent、跨仓或包含多个可独立集成结果的长计划使用 Root → Delivery 两层 Issue Tree。Root 一次批准既定 Delivery，用户只验收 Root。
3.只有不影响产品 / 设计决策的事实查询、机械精确改动，或用户明确要求直接开始 / 无需确认时，才可跳过需求对齐；“为什么流程不一致 / 能力缺失 / 无法编辑 / 应放在哪个入口”不属于纯查询。
4.长计划执行必须读取 .agents/skills/problem-framing/references/long-running-work.md；默认先用一个只读 Scout，Root agent 汇总路径、现状、目标、AC 与集中 Test Batch，并成为唯一 packetizer、assembly owner 和 Control Ledger owner。
5.开发只接收边界明确的 Work Packet；同一 Delivery 可连续复用开发上下文，新 Delivery 使用最小 handoff。开发、reviewer、QA agent 不再嵌套调度 agent。
6.Root 下全部开发与 fixture Work Packet 装配到隔离 assembly 后，只启动一个 fresh QA 做集中测试；不做 per-packet / per-Delivery QA。无依赖、无共享写 / contract / migration / 端口 / 构建冲突的开发 Packet 才可并行，状态与资源检查保持事件驱动。
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.
- 2d ago First seen · 59 lines · 1,444 tokens per session scan A 737979c1d616
1flowbase AGENTS.md is an instructions file published in the GitHub repository taichuy/1flowbase (259 stars, last pushed 3d ago), licensed Apache-2.0. It adds 1,444 tokens to every session, about $0.0072 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 instructions, from other repositories
ccLoad CLAUDE.md
Instructions for caidaoli/ccLoad, covering claude.md, 命令, 测试策略, 代码规范(硬约束) and 架构与入口.
tiygate AGENTS.md
Instructions for tiylabs/tiygate, covering repository guidelines, project structure & module organization, engineering principles, build, test, and development commands and coding style & naming conventions.
10router CLAUDE.md
Instructions for techysy/10router, covering claude.md, what this is, commands, architecture and request flow (the thing to understand first).
unified-ai-system AGENTS.md
Instructions for happy520ai/unified-ai-system, covering repository guidance, ownership, language and module policy, public repository rules and safety.
deltallm AGENTS.md
Instructions for deltawi/deltallm, covering repository instructions, required first step and precedence and maintenance.
yallmap CLAUDE.md
Instructions for kckempf/yallmap, covering llm gateway project context, what this is, architectural decisions already made, v0.1 scope (this session) and gotchas to design for.