Meka AI 工作流 是 AI Skill Hub 本期精选Agent工作流之一。综合评分 7.5 分,整体质量较高。我们推荐使用将其纳入你的 AI 工具库,帮助提升工作效率。
Meka AI 工作流 是一套完整的 AI Agent 自动化工作流方案。通过可视化的节点编排,将复杂的多步骤任务拆解为清晰的自动化流程,实现全程无人值守的智能处理。支持与数百种外部服务和 API 无缝集成,适合构建数据处理管线、业务自动化和 AI 辅助决策系统。
Meka AI 工作流 是一套完整的 AI Agent 自动化工作流方案。通过可视化的节点编排,将复杂的多步骤任务拆解为清晰的自动化流程,实现全程无人值守的智能处理。支持与数百种外部服务和 API 无缝集成,适合构建数据处理管线、业务自动化和 AI 辅助决策系统。
# 方式一:cargo install(推荐) cargo install meka # 方式二:从源码编译 git clone https://github.com/k4yt3x/meka cd meka cargo build --release # 二进制在 ./target/release/meka
# 查看帮助 meka --help # 基本运行 meka [options] <input> # 详细使用说明请查阅文档 # https://github.com/k4yt3x/meka
# meka 配置说明 # 查看配置选项 meka --config-example > config.yml # 常见配置项 # output_dir: ./output # log_level: info # workers: 4 # 环境变量(覆盖配置文件) export MEKA_CONFIG="/path/to/config.yml"
A general-purpose AI agent harness.
[!CAUTION] Agents can perform potentially destructive actions. Exercise caution when granting write permissions. It is not recommended to run meka on important systems with write permissions enabled.
[!WARNING] Agents can consume a large number of tokens on complex tasks. If you're on a subscription, be prepared for quota exhaustion; if you are billed per token, it is recommended that you set a spending limit on the API key.
meka is a general-purpose AI agent harness: it wraps a large language model with a tool set, working memory, persistent sessions, a permission model, and several front-ends (an interactive REPL, one-shot commands, an editor agent via ACP, and an HTTP service). Bring a provider (Claude or OpenAI, API key or subscription) and meka turns it into an agent that reads and edits files, runs commands, searches the web, calls MCP servers, and delegates to sub-agents to accomplish real tasks.
Supported providers:
~/.config/meka/skills/Download a pre-built binary from GitHub Releases, or install with Cargo:
cargo install --locked --git https://github.com/k4yt3x/meka.git
Add a provider profile with meka provider add. It runs the OAuth login (or prompts for an API key), stores the secret in the database, and writes the profile to ~/.config/meka/config.toml:
meka provider add work --type claude-oauth --model claude-opus-4-6
The profile pins a backend type (openai-api, openai-codex, claude-api, or claude-oauth) plus a model. Add several profiles and switch with meka provider use <name> or --provider <name>. For an OpenAI-compatible endpoint like OpenRouter, set --base-url:
meka provider add openrouter --type openai-api --model anthropic/claude-opus-4.6 \
--base-url https://openrouter.ai/api/v1
Run meka and start typing. Press Shift+Tab to cycle permissions (none, read, ask, write):
meka [r] > find all TODO comments in this project
meka [w] > install and start nginx
See the documentation for the full usage guide.
The same agent core is available through several interfaces:
高质量的AI工作流工具,值得关注
AI Skill Hub 为第三方内容聚合平台,本页面信息基于公开数据整理,不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后,再部署至生产环境,并做好必要的安全评估。
✅ MIT 协议 — 最宽松的开源协议之一,可自由商用、修改、分发,仅需保留版权声明。
经综合评估,Meka AI 工作流 在Agent工作流赛道中表现稳健,质量良好。如果你已有明确的使用需求,可以直接上手体验;如果还在评估阶段,建议对比同类工具后再做决策。
| 原始名称 | meka |
| 原始描述 | 开源AI工作流:A general-purpose AI agent harness.。⭐30 · Rust |
| Topics | AI工作流Rust |
| GitHub | https://github.com/k4yt3x/meka |
| License | MIT |
| 语言 | Rust |
收录时间:2026-05-29 · 更新时间:2026-05-30 · License:MIT · AI Skill Hub 不对第三方内容的准确性作法律背书。
选择 Agent 类型,复制安装指令后粘贴到对应客户端