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/xy200303/spec-coding-mcp/agents-mdgit clone --depth 1 https://github.com/xy200303/spec-coding-mcpWrote 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/instructions/xy200303/spec-coding-mcp/agents-md)<a href="https://agentmods.dev/instructions/xy200303/spec-coding-mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/xy200303/spec-coding-mcp/agents-md.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 | $0.00819 | $0.00819 |
| Opus 5 | $0.00409 | $0.00409 |
| Sonnet 5 | $0.00164 | $0.00164 |
| Haiku 4.5 | $0.00082 | $0.00082 |
Grade A, and why
agents 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 3d 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
AGENTS.md
Project: Spec Coding MCP
项目约定(所有人适用)
- 改动小而聚焦,不做无关重构
- 不要猜测业务规则,不确定时先确认
- 不要覆盖他人已做的修改
- 提交前验证
Spec Coding MCP(可选增强)
本项目支持 spec-coding-mcp 工具提供增强开发流程。
检测方法:查看可用工具列表中是否有 spec_context。
◆ 如果你有 spec_context 工具
按以下硬性协议执行:
- 写任何代码前,先调用
spec_context获取当前任务上下文。 - 只执行 spec/TODO 中列出的任务,不从对话历史推导。
- 按清单自上而下执行 →
spec_checkpoint记录进度 → 验证 →spec_done归档。 - UI/UX 任务使用
ui-ux-pro-maxskill。 - 需要原则指导时调用
spec_guidance_read。
常见错误(不要做):不调 spec_context 就写代码 / 跳过验证直接 done / 监听型 TODO 自己标完成。
◆ 如果你没有 spec_context 工具
跳过本节。按项目常规开发流程工作:读需求 → 写代码 → 测试 → 提交。 不要尝试调用 spec_context、spec_checkpoint、spec_done 等不存在的工具。
Guidance 索引(仅 MCP 用户)
需要原则时调用 spec_guidance_read:
| Name | 用途 | 何时读 |
|---|---|---|
engineering |
工程规范、业务确认 | 实现/重构 |
ui-ux |
UI/UX skill 路由 | 前端/页面 |
spec-writing |
spec 工作流、验证器 | 创建/执行 spec |
git-commit |
安全提交流程 | 用户要求 commit |
pr-submit |
PR 创建流程 | 用户要求 PR |
quality-review |
质量自检清单 | done/commit/PR 前 |
Hard Stop(所有人适用)
Ask the user before implementing unclear or high-risk business rules involving money, permissions, state machines, concurrency, idempotency, retries, rollback, compliance, or role differences.
Boundaries(所有人适用)
- Do not guess business rules.
- Do not use stale conversation context over selected specs or open TODOs.
- Do not overwrite user edits or make unrelated reshuffles.
- Keep changes small, focused, and verified.
Shared Memory / Vault(仅 MCP 用户)
specs/ 目录作为持久化 Vault,在不同会话之间共享上下文:
specs/
├── README.md # vault 索引与约定
├── active/ # 当前工作进行中
├── todo/ # 轻量执行清单
├── done/ # 已完成 spec + 最终行为契约
├── review/ # 源码审查任务
└── guidance/ # 可编辑 Agent 指导性提示词
无 MCP 工具时,specs/ 目录退化为普通项目文档目录。
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.
- 3d ago First seen · 95 lines · 819 tokens per session scan A 52b76b6a4a7a
agents is an instructions file published in the GitHub repository xy200303/spec-coding-mcp (2 stars, last pushed 2mo ago), licensed MIT. It adds 819 tokens to every session, about $0.0041 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 2026-08-31.
Other instructions, from other repositories
easy-vibe CLAUDE.md
Claude Code instructions for datawhalechina/easy-vibe, covering claude.md, project overview, development commands, start local documentation server and build/run commands.
easy-vibe AGENTS.md
AGENTS.md instructions for datawhalechina/easy-vibe, covering repository guidelines, project structure & module organization, build, test, and development commands, coding style & naming conventions and testing guidelines.
dotdog AGENTS.md
Instructions for specdog/dotdog, covering agents.md — dotdog, quick start (read this first), commands (when modifying dotdog source) and key rules.
dotdog CLAUDE.md
Instructions for specdog/dotdog, covering claude.md — dotdog, before you do anything, key rules, commands and mcp tools.
vibe-coding-template AGENTS.md
AGENTS.md instructions for humanstack/vibe-coding-template, covering vibe coding template - agent instructions, architecture overview, development standards, code style and architecture patterns.
aider-desk AGENTS.md
AGENTS.md instructions for hotovo/aider-desk, covering agents.md, common commands, development, type checking and linting and formatting.