convez.ai
MCP ServerDesenvolvimento WebAvançado
emirhankudun-ux

SEIS - Ecossistema AI-Native Apple-First

emirhankudun-ux/SEIS

Ecossistema AI-native completo para agentes, MCP e produtos digitais sustentáveis

Ver no GitHub
BomPluginScore · de 0 a 100

Plataforma open source abrangente para construir, governar e evoluir sistemas de IA, agentes, MCP servers, skills, LLM workflows e produtos full-stack com foco em arquitetura sustentável. Cobre IA, engenharia, dados/ML, design e open source governance, priorizando Apple mas mantendo suporte forte em Windows, Android, Web e infraestrutura.

Instalação

$npx convez-ai emirhankudun-ux/SEIS

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
1
Forks
0
Acompanhando
1
Saúde
100%
Último commit
há 1 semana
Versão
—
Issues abertas
101
Contribuidores
3

Casos de uso

  • Desenvolvimento de agentes IA e orquestração com MCP servers
  • Governança e arquitetura de sistemas AI-native de longa duração
  • Construção de produtos full-stack com design systems integrados
  • Pesquisa e implementação de ML, NLP e computer vision
  • Engenharia de dados, RAG e avaliação de modelos

Veredito

SEIS é um projeto ambicioso e bem-estruturado que vai além de um simples repositório, funcionando como plataforma-camada para sistemas AI. A ênfase em governança clara, sustentabilidade arquitetural e segurança o diferencia de abordagens mais caóticas, mas exige senioridade técnica significativa para aproveitar todo seu potencial.

README

A tradução para o português deste README ainda está na fila.

SEIS

SEIS is an AI-native open source platform for building, governing, and evolving software, design, data, automation, and agent systems as one sustainable digital ecosystem.

It is not just an application repository. SEIS is a living platform layer for AI systems, AI agents, MCP servers, skills, plugins, LLM workflows, full-stack products, design systems, data systems, and long-term engineering governance.

Mission

SEIS aims to become one of the most comprehensive open source AI-native development ecosystems on GitHub: useful for engineers, designers, AI researchers, educators, maintainers, and builders who need a clear operating model for modern digital products.

The repository optimizes for:

  • maintainable architecture over shortcuts
  • security, reliability, and sustainability over feature volume
  • clear governance over branch chaos
  • high-quality open source adoption over noisy growth
  • AI-assisted engineering with human review and accountable decisions

Core Scope

AreaSEIS covers
AI systemsAI agents, agent orchestration, MCP, skills, plugins, LLM routing, memory, RAG, model evaluation, AI safety
Engineeringalgorithms, data structures, full stack, frontend, backend, mobile, desktop, game systems, embedded, robotics, compilers, architecture, testing, SRE, DevOps, cloud, cybersecurity
Data and MLdata engineering, big data, ML, deep learning, generative AI, NLP, computer vision, knowledge graphs, governance, ethics
Designproduct design, design systems, UX engineering, UI engineering, interaction, motion, branding, typography, accessibility, calm technology
Open sourcemain-centered governance, contribution quality, security reporting, documentation, discoverability, community growth

Repository Model

main is the only permanent branch for SEIS.

Temporary working branches may exist only as review or integration staging. Long-term development must merge back into main, and branch cleanup should be explicit, reviewed, and non-destructive. Do not delete local or remote branches only because they look stale; first verify merge state, ownership, and rollback needs.

Platform Strategy

SEIS is broad, but not careless. It is Apple-first whenever practical while remaining strong across Windows, Android, Web, AI, data, and infrastructure. Languages and frameworks are included when they strengthen a platform lane or a quality gate.

PlatformPriority languages
Apple firstSwift, SwiftUI, Objective-C, Metal, AppKit, UIKit, Combine, Core Data, CloudKit
WindowsC#, .NET, C++, Rust, WinUI, TypeScript, Go, Zig, Python when needed
AndroidKotlin, Java, Jetpack Compose, C++, Rust, TypeScript, Go, Python when needed
Web and AI toolingTypeScript, HTML, CSS, Go, Rust, Python when needed, JavaScript when needed

Unused SDKs, runtimes, and language toolchains are not installed by default. Local development should stay fast, simple, and reversible. CI may install specialized tools only when a specific check requires them.

Source Languages and Ecosystem Stack

SEIS keeps the GitHub language surface honest: real source languages are counted from real files, while frameworks, SDKs, cloud products, databases, IDEs, design tools, and productivity systems live in a separate ecosystem stack.

SurfaceCanonical recordRule
GitHub source languagesreports/language-distribution.mdCount real source languages only and track platform-family balance without filler code.
Full-stack language matrixreports/fullstack-language-matrix.mdRoute source/config languages by platform layer.
SEIS ecosystem stackreports/seis-technology-stack.mdShow frameworks, tools, clouds, databases, and design systems outside the language bar.
SEIS AGI systemreports/seis-agi-system.mdGovern agent systems, memory, planning, research automation, MCP, skills, plugins, token efficiency, and the 90-day release window.

The long-term GitHub language profile should signal a real multi-platform ecosystem: Apple/Swift 25-30%, AI/Data/Python/SQL 18-22%, TypeScript/JavaScript tooling 15-20%, Android/JVM 10-15%, Rust/C/C++ systems 10-15%, Go/Infrastructure 5-8%, Windows/.NET 5-8%, and HTML/CSS previews 0-3%. Do not add filler code to change percentages.

Run npm run check:seis-technology-stack after changing stack categories and npm run check:seis-agi-system after changing agent, memory, planning, research, MCP, skills, plugin, or token-efficiency policy.

AGI System and Three-Month Roadmap

SEIS now tracks its AGI direction as a human-owned, Apple-first assistant architecture: advanced agents, memory systems, planning systems, research automation, MCP, skills, plugins, data, design, development, and interactive read/write workflows are routed through explicit safety gates.

SurfaceCanonical recordRule
AGI system contractreports/seis-agi-system.mdTrack the 150-domain capability atlas, plugin/MCP lanes, token efficiency target, and release window.
Active mission boardreports/seis-active-mission-board.mdConvert the three-month goal into Month 1/2/3 execution lanes with acceptance gates and evidence paths.
Apple-first Swift contractpackages/seis_platform_swift/Sources/SeisPlatformKit/SeisAGISystemContract.swiftKeep the implementation path tied to Swift, SwiftUI, Objective-C, Metal, AppKit, UIKit, Combine, Core Data, and CloudKit.
Generated implementation docdocs/agi/seis-agi-system.mdExplain implementation roots, token efficiency, plugin use, and visual source policy.

Architecture Map

PathPurpose
packages/seis-aiAI agent CLI, MCP server, audit tools, prompts, resources, and tests
mcpSEIS MCP server entrypoints and integration surface
plugins/seisCodex plugin bundle, scripts, and SEIS skill entrypoints
plugins/seis-ai-agentSEIS-Agent unified orchestration across cloud, code, design, data, memory, context, MCP, skills, plugins, and automation
plugins/seis-cloudDedicated SEIS Cloud deployment, server-target, and cloud-readiness plugin package
plugins/seis-codeDedicated SEIS-Code engineering plugin package
plugins/seis-designDedicated SEIS-Design product and design-system plugin package
plugins/seis-dataDedicated SEIS-DATA analytics and knowledge-governance plugin package
packages/seis_platform_swiftApple platform policy package
packages/seis_windows_csharpWindows platform policy package
packages/seis_kernelCapability, language, plugin, platform, and AGI-system contract builders
packages/seis_kernel_goGo governance and readiness policy contracts
polyglotCross-language audit lanes and platform proof-of-concept surfaces
apps/webBrowser-facing product and documentation surface
apps/androidAndroid direction and validation notes
apps/macosmacOS direction and Apple-native notes
docsArchitecture, governance, deployment, strategy, quality, and research records
reportsGenerated ecosystem, language, capability, and readiness reports

AI-Assisted Development

SEIS is built with human ownership and AI assistance.

  • Maintainer: Emirhan Kudun (@emirhankudun-ux)
  • Primary execution assistant: OpenAI Codex / ChatGPT
  • Architecture and review assistant: Claude

Technology names in this repository may appear as platform references, compatibility targets, or tool integrations. They do not imply endorsement, employment, sponsorship, or direct contribution unless explicitly stated by the maintainer.

Quality Gates

Use the lightest reliable check first, then scale validation with risk:

npm run check:open-source-governance
npm run seis:check
npm run check:seis-platform-language-policy
npm run check:seis-platform-kernel
npm run check:seis-active-mission-board

No command above installs a new local language runtime. Some checks skip optional toolchains when they are not present.

GitHub Actions also runs CodeQL code scanning for JavaScript, TypeScript, and Python on relevant pull requests, main pushes, weekly scheduled scans, and manual dispatches. This keeps security scanning in GitHub without asking local contributors to install every platform SDK.

Contribution Path

Start with:

Good contributions are small, reviewed, reversible, documented, and aligned with the long-term platform model.

GitHub Growth Strategy

SEIS should be discoverable, useful, and trustworthy before it tries to be loud. The GitHub surface should continuously improve through:

  • clear README, repository metadata, and topic positioning
  • strong issue and PR templates
  • structured Discussions for ideas, Q&A, and show-and-tell
  • a visible support path that keeps questions, bugs, features, and security reports in the right place
  • release notes and GitHub Pages links that explain what changed
  • accurate contributor attribution
  • quality-first examples and demos
  • security and governance transparency
  • approachable docs for agents, MCP, skills, plugins, and platform lanes

Use issues for actionable bugs and scoped work. Use Discussions for architecture questions, broad ideas, and examples that need community shaping before they become work items.

The GitHub adoption model is tracked in docs/governance/github-market-readiness.md.

Repository Metadata

The public GitHub repository should describe SEIS as an AI-native, Apple-first, full-stack, design-driven open source ecosystem. Topics should reflect the real architecture: AI agents, MCP, LLM workflows, Swift, SwiftUI, TypeScript, Rust, design systems, DevOps, security, and GitHub Pages.

Current Status

SEIS is actively evolving. Some historical documents still record earlier private or consolidation phases. The current direction is open source, main-centered, AI-native, security-conscious, and architecture-led.

Ver o original no GitHub

Bom para

Relacionados