CCM
/Skills
SkillsMCPMarketplacesDigestToolsAdvertise

This week in Claude

Every Monday: Claude Code, Agent SDK, MCP, and the Anthropic platform moves worth your time.

Skills by Category
Frontend DevelopmentBackend & APIsTesting & QASecurityDevOps & CI/CDGit & Pull RequestsDocumentationCode Review & QualityAI & Agent BuildingSkill Development
MCP Servers by Category
Sales & MarketingWeb & Browser AutomationDatabasesAI & LLM ToolsCloud & InfrastructureCommunication & MessagingDeveloper ToolsDesign & CreativeDocuments & KnowledgeSearch & Web Crawling
Marketplaces by Category
AI Agents & OrchestrationLLM IntegrationDevelopment ToolsFrontend & UIBackend & APIsDatabasesTesting & Code QualityDevOps & CloudSecurity & ComplianceGit & Version Control

Claude Code Marketplaces

Discover Claude Code plugins, extensions, and tools. Automatically updated directory of Anthropic Claude AI marketplaces with development tools, productivity plugins, and integrations.

Resources

  • Browse Skills
  • Browse MCP Servers
  • Browse Marketplaces
  • Skill index
  • MCP index
  • Marketplace index
  • Plugins Reference

Community

  • About
  • Tools
  • Feedback
  • Privacy Policy
  • Advertise

Built for the Claude Code community with Claude Code by mertbuilds.com

Independent project, not affiliated with Anthropic
alchaincyf avatar

Huashu Agent Swarm

alchaincyf/huashu-skills
625 installs1.4k stars
Summary

This spins up multiple Claude instances that work in parallel on the same codebase, coordinating through git alone with no master orchestrator. Each agent claims tasks via lock files, commits independently, and pulls others' work in an infinite loop. Inspired by Nicholas Carlini's compiler project. You get a tmux session with 8 agents by default, a web dashboard to watch their progress, and git worktrees to keep them isolated without cloning repeatedly. It's genuinely experimental infrastructure for tackling large projects when you're willing to babysit the chaos. Watch for API rate limits and merge conflicts, but the self-organizing approach through git is legitimately interesting if you want to see what emerges from parallel AI collaboration.

Install to Claude Code

npx -y skills add alchaincyf/huashu-skills --skill huashu-agent-swarm --agent claude-code

Installs into .claude/skills of the current project.

CodeRabbit
CodeRabbit
AI writes the code. CodeRabbit catches the slop.
Try For Free →
AppSignal
AppSignal
Monitor with ease. Code with confidence.
Start Free Trial →
Make money from your Skills
Make money from your Skills
On Capafy, your Skill runs online 24/7 as an agent product, and you get paid every time someone uses it.
Start earning →
Put your SEO on autopilot
Put your SEO on autopilot
An agent that runs the SEO playbooks that move rankings and ships PRs you control.
Get founding access →
Vibe Prospecting MCPVibe Prospecting MCP
Vibe Prospecting MCP
Connect Claude to +800M contacts, +150M companies. Find & Enrich leads in chat.
Try For Free →
Make your agent a DeFi expert
Make your agent a DeFi expert
Agent, run crypto. Access onchain data & trade routes via 1inch.
Install now →
CodeRabbit
CodeRabbit
AI writes the code. CodeRabbit catches the slop.
Try For Free →
AppSignal
AppSignal
Monitor with ease. Code with confidence.
Start Free Trial →
Make money from your Skills
Make money from your Skills
On Capafy, your Skill runs online 24/7 as an agent product, and you get paid every time someone uses it.
Start earning →
Put your SEO on autopilot
Put your SEO on autopilot
An agent that runs the SEO playbooks that move rankings and ships PRs you control.
Get founding access →
Vibe Prospecting MCPVibe Prospecting MCP
Vibe Prospecting MCP
Connect Claude to +800M contacts, +150M companies. Find & Enrich leads in chat.
Try For Free →
Make your agent a DeFi expert
Make your agent a DeFi expert
Agent, run crypto. Access onchain data & trade routes via 1inch.
Install now →
Files
SKILL.mdView on GitHub

Infinite Agent Loop - 无限Agent蜂群模式

受Nicholas Carlini用16个Claude实例自主构建C编译器的启发。 没有master agent,纯git自组织,每个agent独立认领任务、写代码、推送。

触发条件

当用户提到「蜂群模式」「多agent并行」「infinite loop」「agent swarm」「启动蜂群」时使用此技能。

前置要求

  • tmux(brew install tmux)
  • claude CLI(已安装)
  • git 仓库(已有或新建)

使用流程

Step 1: 描述项目

用户告诉我:

  • 项目目录路径(必须是git仓库)
  • 项目目标和总体描述
  • 初始任务列表(或让agent自行拆解)
  • agent数量(默认8个)
  • 代码规范和测试命令

Step 2: 初始化项目

bash SKILL_DIR/scripts/setup_project.sh <项目目录>

这会在项目中创建:

  • AGENT_PROMPT.md - 从模板生成,需要我根据用户需求定制
  • TASKS.md - 初始任务清单
  • current_tasks/ - 任务认领目录
  • agent_logs/ - 日志目录

然后我根据 references/agent-prompt-template.md 定制 AGENT_PROMPT.md,填入项目具体信息。

Step 3: 启动蜂群

bash SKILL_DIR/scripts/start_swarm.sh <agent数量> <项目目录>

这会:

  1. 为每个agent创建 git worktree(共享.git对象库,不浪费磁盘)
  2. 创建 tmux session,每个pane一个agent
  3. 每个agent进入无限循环:pull → 认领任务 → 执行 → push → 下一个

Step 4: 打开观测台

python3 SKILL_DIR/scripts/dashboard.py <项目目录> 8420

浏览器打开 http://localhost:8420,可以:

  • 实时查看所有agent状态、git log、任务进度
  • 查看每个agent的最新日志
  • 输入框直接发送指令给agent(写入HUMAN_INPUT.md)
  • 一键停止所有agent

也可以用命令行监控:

# 终端状态
bash SKILL_DIR/scripts/status.sh <项目目录>

# 发送指令
bash SKILL_DIR/scripts/send_input.sh <项目目录> "你的指令"

# 直接进入tmux观察
tmux attach -t swarm-<项目名>

Step 5: 停止

bash SKILL_DIR/scripts/stop_swarm.sh <项目目录>

自动停止所有agent + 合并分支 + 清理worktrees。

核心机制

Git自组织协调

  • 每个agent通过 current_tasks/*.lock 文件认领任务
  • 通过 TASKS.md 了解全局进度
  • 通过 git log 了解其他agent的工作
  • 冲突由agent自行解决

Git Worktree隔离

  • 不用多份clone,用 git worktree 实现隔离
  • 所有worktree共享同一个 .git 对象库
  • 每个agent在自己的worktree独立工作

无限循环

  • 每个agent完成一个session后自动开始下一个
  • 通过 git pull 获取其他agent的最新成果
  • 通过 sleep 间隔避免API限流

关键配置

参数默认值说明
agent数量8可在启动时指定
sleep间隔5秒agent_loop.sh中可调
模型claude-opus-4-6agent_loop.sh中可调

风险和应对

风险应对
API限流sleep间隔 + 可调agent数量
合并冲突AGENT_PROMPT指导小粒度commit
死循环无用功日志监控 + 停止条件
磁盘空间stop_swarm.sh自动清理
成本失控可在AGENT_PROMPT中限制session数

花叔出品 | AI Native Coder · 独立开发者 公众号「花叔」| 30万+粉丝 | AI工具与效率提升 代表作:小猫补光灯(AppStore付费榜Top1)·《一本书玩转DeepSeek》

Featured
CodeRabbit
CodeRabbit
AI writes the code. CodeRabbit catches the slop.
Try For Free →
AppSignal
AppSignal
Monitor with ease. Code with confidence.
Start Free Trial →
Make money from your Skills
Make money from your Skills
On Capafy, your Skill runs online 24/7 as an agent product, and you get paid every time someone uses it.
Start earning →
Put your SEO on autopilot
Put your SEO on autopilot
An agent that runs the SEO playbooks that move rankings and ships PRs you control.
Get founding access →
Vibe Prospecting MCPVibe Prospecting MCP
Vibe Prospecting MCP
Connect Claude to +800M contacts, +150M companies. Find & Enrich leads in chat.
Try For Free →
Make your agent a DeFi expert
Make your agent a DeFi expert
Agent, run crypto. Access onchain data & trade routes via 1inch.
Install now →
Categories
Git & Pull RequestsAI & Agent Building
First SeenJun 3, 2026
View on GitHub

More from alchaincyf/huashu-skills

All 21 skills →
  • Huashu Prompt Save617
  • Huashu Md To Pdf990
  • Huashu Douyin Script922
  • Huashu Proofreading813
  • Huashu Slides777
  • Huashu Design716
  • Huashu Wechat Image694
  • Huashu Data Pro692
  • Huashu Research686
  • Huashu Xhs Image686
  • Huashu Script Polish685
  • Huashu Topic Gen677
  • Huashu Speech Coach663
  • Huashu Article Edit649
  • Huashu Video Outline647
  • Huashu Article To X645
  • Huashu Info Search645
  • Huashu Image Upload634
  • Huashu Material Search628
  • Huashu Video Check626

Recommended

More Git & Pull Requests →
googleworkspace avatar
recipe-audit-external-sharing

googleworkspace/cli

Find and review Google Drive files shared outside the organization.
624
30.4k
seflless avatar
deepwiki

seflless/deepwiki

Query any public GitHub repo's documentation via DeepWiki. Use when needing to understand a library, framework, or dependency. Triggers on "look up docs", "how does X work", "deepwiki", "deepwiki".
623
21
alirezarezvani avatar
copy-editing

alirezarezvani/claude-skills

When the user wants to edit, review, or improve existing marketing copy. Also use when the user mentions 'edit this copy,' 'review my copy,' 'copy feedback,' 'proofread,' 'polish this,' 'make this better,' or 'copy sweep.' This skill provides a systematic approach to editing marketing copy through multiple focused passes.
621
24.6k
mims-harvard avatar
tooluniverse-literature-deep-research

mims-harvard/tooluniverse

Deep literature review — PubMed, EuropePMC, bioRxiv preprints, citation networks, evidence synthesis. Disambiguates queries, runs collision-aware searches, grades evidence T1-T4, and produces structured reports. Use for systematic literature review, meta-analysis evidence collection, and detailed answer-with-citations workflows.
620
1.6k
alirezarezvani avatar
seo-audit

alirezarezvani/claude-skills

When the user wants to audit, review, or diagnose SEO issues on their site. Also use when the user mentions "SEO audit," "technical SEO," "why am I not ranking," "SEO issues," "on-page SEO," "meta tags review," or "SEO health check." For building pages at scale to target keywords, see programmatic-seo. For adding structured data, see schema-markup.
619
24.6k
alirezarezvani avatar
competitive-teardown

alirezarezvani/claude-skills

Analyzes competitor products and companies by synthesizing data from pricing pages, app store reviews, job postings, SEO signals, and social media into structured competitive intelligence. Produces feature comparison matrices scored across 12 dimensions, SWOT analyses, positioning maps, UX audits, pricing model breakdowns, action item roadmaps, and stakeholder presentation templates. Use when conducting competitor analysis, comparing products against competitors, researching the competitive landscape, building battle cards for sales, preparing for a product strategy or roadmap session, responding to a competitor's new feature or pricing change, or performing a quarterly competitive review.
615
24.6k