artemis-code
Advanced tools
Local-first AI engineering, visual generation, long-video production, memory, tools, and mobile automation in one command-line workspace.
Local-first AI engineering, visual generation, long-video production, memory, tools, and mobile automation — in one command-line workspace.
Created by 420.COMPANY · npm: artemis-code · GitHub: 420company/artemis
Artemis Code is a local-first AI workspace agent for people who want execution, not suggestions.
It works inside your real project folder, reads the files, edits the code, runs commands, checks logs, verifies results, and keeps context across long sessions. It can build software, review changes, manage tools, generate images and video, operate long creative workflows, remember your preferences, and connect to mobile chat platforms so you can drive work from anywhere.
Artemis is designed for users who want one intelligent operator across engineering, creative production, research, automation, and daily workflows.
Current npm release: 0.2.72
Artemis runs where your work lives: your terminal, your repository, your files, your machine.
You can ask for a feature, a fix, a release check, a refactor, a README rewrite, a visual asset, or a long video. Artemis will inspect the workspace, plan the steps, make the changes, run the appropriate checks, and report the result with evidence.
What this means in practice:
Artemis can handle everyday and advanced engineering work:
The goal is simple: you describe the outcome; Artemis does the operational work.
Long work often fails because the assistant forgets. Artemis is built to preserve continuity.
It maintains local memory, session state, collapse ledgers, tool evidence, and recovery context so long tasks can continue without losing the important parts. Your preferences, project conventions, workflow habits, and recurring constraints can become part of the way Artemis works with you.
Useful for:
Artemis includes a full visual workflow layer for image and video generation.
It can help create:
The visual system is provider-aware and can route image, vision, and video requests through configured providers. It asks for missing inputs, handles local media references, and keeps creative intent connected to the final generation workflow.
Saga is Artemis' long-video production workflow. It is built for videos that need structure, consistency, and continuity instead of one-off clips.
Saga can:
Saga is designed for users who want to say: “Make this into a real video,” then be guided through the right creative and technical steps.
Artemis can work with both complete scripts and partial inspiration.
If you already have a script, Artemis treats it as the controlling narrative. If you only have a topic, mood, character, place, or rough idea, Artemis can act as a screenwriter and expand it into a structured Saga-compatible brief.
The package includes a full bilingual authoring guide:
docs/saga-brief-authoring-guide.htmlIt explains how to write timecoded scenes, dialogue markers, aspect ratio notes, character locks, opening framing, world anchors, audio intent, BGM planning, and advanced long-video brief structures.
Artemis can be connected to chat platforms such as Telegram, Discord, or WeChat through its bridge system.
This lets you:
The mobile bridge turns your local machine into a reachable creative and engineering workstation.
Artemis is not limited to code. It includes workflow tools for everyday operations and can connect to external services through MCP servers.
Capabilities include:
The intention is to make Artemis useful as a practical daily operator, not only a coding assistant.
For long-running work, Artemis can detach tasks into background workflows. This is useful for large investigations, deep refactors, extended research, or slow media operations.
You can keep using your terminal while Artemis continues the heavy work and returns when there is a result.
Requirements:
Install globally:
npm install -g artemis-code
Start Artemis inside any project:
cd /path/to/your/project
artemis
Fix the failing build and run the tests.
Review my current Git diff for release blockers.
Create a polished product image for this landing page.
Turn this story idea into a 60-second cinematic Saga video.
Clean the package, verify it, bump the version, and publish to npm.
Rewrite the README for GitHub so it explains the product clearly to users.
/config — Configure providers, models, keys, and preferences/team — Let Artemis choose the best routing strategy for the task/review — Review the current Git diff and identify risks/nidhogg — Run heavy or long work in the background/wordup — Save important context into memory/soul — Define long-term personal style, rules, and working preferences/mcp — Manage external MCP integrationsArtemis is for builders, founders, creators, designers, engineers, and operators who want a single local agent that can actually do the work.
Use it when you want:
Artemis Code 是一个本地优先的 AI 工作区代理。它不是只给建议、让你自己复制粘贴的聊天框，而是可以直接进入你的真实项目目录，读取文件、修改代码、执行命令、检查日志、验证结果，并在长任务中保持上下文连续的执行型助手。
它可以写代码、修 Bug、做发布检查、清理包内容、生成图片和视频、组织长视频工作流、记住你的偏好，也可以通过手机聊天平台远程接收指令和发送产物。
Artemis 面向的是希望把工程、创意、自动化、研究和日常操作交给一个统一智能操作者的人。
当前 npm 版本：0.2.72
Artemis 运行在你的终端和项目目录里。它面对的不是抽象问题，而是真实文件、真实命令、真实构建、真实错误。
你可以让它做一个功能、修一个问题、检查一次发布、重写文档、生成视觉素材，或者制作一段长视频。Artemis 会自己检查工作区、拆解步骤、修改文件、运行验证，并基于工具结果汇报进展。
实际效果是：
Artemis 可以处理日常和复杂的软件工程任务：
你描述目标，Artemis 负责执行过程。
很多 AI 工具在长任务中会遗忘前文。Artemis 的设计目标是让任务可以持续推进。
它会把记忆、会话状态、工具证据、压缩账本和恢复上下文保存在本地，让长时间工作不会因为中断、折叠或会话变长而失去关键线索。你的偏好、项目规则、语言风格和长期约束也可以被保留下来。
适合用于：
Artemis 内置完整的视觉工作流，可以处理图片、视觉分析和视频生成。
它可以帮助你创建：
视觉系统会根据配置的模型和供应商选择合适路径，主动询问缺失参数，识别本地媒体引用，并把创意意图稳定传递到最终生成流程。
Saga 是 Artemis 的长视频生产工作流。它不是简单生成一个短片段，而是为需要结构、连续性和可控性的完整视频而设计。
Saga 可以：
Saga 适合用户直接说：“把这个想法做成一条真正的视频。”然后由 Artemis 引导完成创作和技术流程。
Artemis 可以处理完整剧本，也可以处理只有一句话的灵感。
如果你已经有剧本，Artemis 会把它作为权威叙事来执行。如果你只有主题、氛围、人物、地点或大致想法，Artemis 可以进入 AI 编剧模式，把它扩展成 Saga 能稳定识别的结构化 brief。
包内包含完整中英文说明书：
docs/saga-brief-authoring-guide.html它详细说明了时间码、对白标记、画幅比例、角色锁定、首帧定位、世界锚点、音频意图、BGM 规划和高级长视频 brief 写法。
Artemis 可以通过桥接系统连接 Telegram、Discord 或微信等聊天平台。
你可以：
这让 Artemis 不只是终端工具，也可以成为随时可用的 ambient agent。
Artemis 不只处理代码。它也可以接入日常工具和外部服务。
能力包括：
目标是让 Artemis 成为一个实用的日常操作者，而不仅仅是编程助手。
对于耗时任务，Artemis 可以把工作转入后台执行。适合大型排查、深度重构、长时间研究和媒体处理。
你可以继续使用终端，Artemis 在后台推进任务，并在有结果后回来汇报。
环境要求：
全局安装：
npm install -g artemis-code
进入任意项目目录并启动：
cd /path/to/your/project
artemis
修复现在失败的构建，并跑完测试。
检查我当前的 Git diff，找出发布风险。
给这个落地页生成一张高级产品视觉图。
把这个故事想法扩展成 60 秒 Saga 电影感视频。
清理 npm 包内容，验证、升级版本并发布。
重写 GitHub README，让用户一眼看懂产品能力。
/config — 配置模型供应商、密钥和偏好/team — 让 Artemis 自动选择最合适的任务路线/review — 审查当前 Git diff，发现潜在风险/nidhogg — 把复杂或耗时任务转入后台执行/wordup — 保存重要上下文到记忆/soul — 定义长期个人风格、规则和工作偏好/mcp — 管理外部 MCP 集成Artemis 适合开发者、创作者、设计师、创业者、运营者，以及任何希望用一个本地 AI 代理真正完成工作的人。
当你需要这些能力时，Artemis 会特别有用：
FAQs
Local-first AI engineering, visual generation, long-video production, memory, tools, and mobile automation in one command-line workspace.
We found that artemis-code demonstrated a healthy version release cadence and project activity because the last version was released less than a year ago. It has 1 open source maintainer collaborating on the project.
Did you know?
Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.
Research
/Security News
A mini Shai-Hulud campaign compromised Red Hat Cloud Services npm packages to steal developer and CI/CD secrets during installation.
Research
/Security News
The North Korean malware loader hides in a Packagist-listed package and its GitHub branch to fetch and execute remote code in a likely Contagious Interview-style lure.
Security News
The Rust project is moving toward formal rules on LLM use in contributions after months of internal debate over maintainer burden, code quality, and contributor experience.