S
SkillNav

Cokodo Agent

AI 与智能体

by dinwind

向兼容 MCP 的 IDE 提供 .agent/ 项目上下文、参考资料与协作信息，便于智能开发。

GitHub

什么是 Cokodo Agent？

向兼容 MCP 的 IDE 提供 .agent/ 项目上下文、参考资料与协作信息，便于智能开发。

README

<!-- GitHub repo "About" description (single line): AI collaboration protocol + CLI (cokodo-agent). SDD change units + MCP; Python-native alternative/complement to OpenSpec. -->

AI Agent Collaboration Protocol

Protocol PyPI

What does cokodo-agent do?

cokodo-agent (co) is a CLI + convention directory .agent/ that turns “how AI collaborates, what it remembers, and which rules it codes by” into versionable, syncable, checkable project assets.

ProblemApproach
AI “forgets” every new chatSingle entry point: start-here.md, status.md, MCP for on-demand context
Each IDE has its own rulesco adapt generates Cursor / Claude / Copilot / Gemini / Codex entry files once; the protocol stays in .agent/ only
Spec and implementation driftv1.9+ uses co change to manage “proposal → tasks → archive” under project/changes/, SDD-style like OpenSpec, with co lint checks

In short: Ship protocol + session state + optional SDD change units inside the repo, wired by Python and co serve; it does not replace the IDE, it fixes “project-level context.”

cokodo-agent vs OpenSpec

The two can be used together; cokodo keeps SDD inside .agent/ and is oriented toward “one repo, one toolchain.”

AspectOpenSpeccokodo-agent
FocusAlign on “what to build” before coding (change units + living specs)Align on “who the project is, where it stands” (protocol + cross-session state)
Changes/workflowchanges/ → propose → apply → archive; slash commandsco change new | apply | list | archive; same model as status.md and MCP
Living specsopenspec/specs/ as first-class.agent/project/specs/ + optional merge on archive
MCPOfficially no MCPco serve / co serve --workspace; IDE queries context, status, changes
Protocol upgradeManual template follow-upco diff / co sync; project/ is not overwritten
Multi-IDEAGENTS.md + per-IDE slash commandsco adapt cursor|claude|copilot|gemini|codex|all
RuntimeNode.js 20.19+Python 3.10+
Cross-projectSingle-repo orientedco ref / co collab / global registration

Relationship: OpenSpec excels at spec and change-unit expression and community workflow; cokodo-agent excels at session continuity, MCP, protocol versioning, and one flow for all IDEs. From v1.9, cokodo includes SDD change units; see OpenSpec comparison research and ADR-009 for tradeoffs. Thanks to OpenSpec (Fission-AI) for the inspiration.

Common scenarios and co commands (concise)

Install and version

bash
pipx install cokodo-agent          # or pip install cokodo-agent
# MCP is included; use co serve for IDE integration
co version

New project setup

bash
cd your-repo
co init                    # interactive: create .agent/
co init -y --tools all     # non-interactive + generate all IDE entry files
# then edit .agent/project/context.md, tech-stack.md

Existing .agent/, only add or refresh IDE entry files

bash
co adapt cursor            # or claude / copilot / gemini / codex / all
co detect                  # see which IDE files already exist in the repo

Align protocol with repo .agent/

bash
co diff                    # compare with remote/bundled protocol
co sync -y                 # upgrade core etc., keep project/
co lint                    # compliance + changes-structure (when using change units)

One feature, spec-driven (SDD)

bash
co change new feature-x           # skeleton: proposal / specs / design / tasks
co change apply feature-x        # set as active, update status.md
co change list                    # see progress; when done:
co change archive feature-x --merge-specs   # archive, optionally merge specs snapshot
co change new fix-y --schema minimal        # small change: specs + tasks only

Use MCP from the IDE (less typing, direct protocol access)

bash
co serve                   # stdio; configure in Cursor/Claude/Copilot/Gemini/Codex
co serve --workspace       # serve multiple project directories
# entry files from co adapt include MCP setup notes

Multi-repo context

bash
co ref add ../other/.agent --name other
co collab add ../lib/.agent --name lib --role replica
co ref check / co collab status

Other common commands

bash
co status                  # view or init status.md
co scaffold                # fill missing project/ files
co context                 # list context files by stack/task
co journal                 # append to session-journal
co update-checksums        # maintainers: refresh manifest checksums

Further documentation

DocumentDescription
Usage guide (English)Full command reference
使用指南 (中文)命令与选项最全
Protocol / CLI version configHow config and co version / lint use versions
cokodo-agent/README.mdPackage and development notes
OpenSpec comparison researchPoint-by-point comparison and ADR

Agent Protocol overview (.agent/)

The protocol uses engine / instance separation: core/ holds generic rules, project/ holds project state and context. Removing .agent/ should not affect the build; copy it to another project to reuse.

code
.agent/
├── start-here.md      # AI reads this first
├── manifest.json
├── core/              # Governance engine
├── project/           # context, status, changes, specs…
├── adapters/
└── scripts/

Repository structure

code
agent_protocol/
├── .agent/           # Protocol reference implementation
├── cokodo-agent/     # CLI source (PyPI package)
└── docs/             # Usage guides and SOP

Acknowledgements

  • OpenSpec — Spec-driven development and change-unit workflow; this repo’s v1.9 co change and project/changes/ are inspired by it, implemented with Python and native MCP integration. See .agent/project/research/openspec-analysis.md.

Protocol: 3.2.1 | CLI: see cokodo-agent

常见问题

Cokodo Agent 是什么？

向兼容 MCP 的 IDE 提供 .agent/ 项目上下文、参考资料与协作信息，便于智能开发。

相关 Skills

Claude接口

by anthropics

Universal
热门

面向接入 Claude API、Anthropic SDK 或 Agent SDK 的开发场景，自动识别项目语言并给出对应示例与默认配置，快速搭建 LLM 应用。

✎ 想把Claude能力接进应用或智能体，用claude-api上手快、兼容Anthropic与Agent SDK，集成路径清晰又省心

AI 与智能体
未扫描116.0k

计算机视觉

by alirezarezvani

Universal
热门

聚焦目标检测、图像分割与视觉系统落地，覆盖 YOLO、DETR、Mask R-CNN、SAM 等方案，适合定制数据集训练、推理优化及 ONNX/TensorRT 部署。

✎ 把目标检测、图像分割到推理部署串成完整工程链路，主流框架与 YOLO、DETR、SAM 等方案都覆盖，落地视觉 AI 会省心很多。

AI 与智能体
未扫描10.7k

RAG架构师

by alirezarezvani

Universal
热门

聚焦生产级RAG系统设计与优化，覆盖文档切块、检索链路、索引构建、召回评估等关键环节，适合搭建可扩展、高准确率的知识库问答与检索增强应用。

✎ 面向RAG落地，把知识库、向量检索和生成链路系统串联起来，做架构设计时更清晰，也更少踩坑。

AI 与智能体
未扫描10.7k

相关 MCP Server

知识图谱记忆

编辑精选

by Anthropic

热门

Memory 是一个基于本地知识图谱的持久化记忆系统，让 AI 记住长期上下文。

✎ 帮 AI 和智能体补上“记不住”的短板，用本地知识图谱沉淀长期上下文，连续对话更聪明，数据也更可控。

AI 与智能体
83.6k

顺序思维

编辑精选

by Anthropic

热门

Sequential Thinking 是让 AI 通过动态思维链解决复杂问题的参考服务器。

✎ 这个服务器展示了如何让 Claude 像人类一样逐步推理，适合开发者学习 MCP 的思维链实现。但注意它只是个参考示例，别指望直接用在生产环境里。

AI 与智能体
83.6k

PraisonAI

编辑精选

by mervinpraison

热门

PraisonAI 是一个支持自反思和多 LLM 的低代码 AI 智能体框架。

✎ 如果你需要快速搭建一个能 24/7 运行的 AI 智能体团队来处理复杂任务（比如自动研究或代码生成），PraisonAI 的低代码设计和多平台集成（如 Telegram）让它上手极快。但作为非官方项目，它的生态成熟度可能不如 LangChain 等主流框架，适合愿意尝鲜的开发者。

AI 与智能体
6.9k

评论