Skip to content
View deb888's full-sized avatar
🏠
Working from home
🏠
Working from home

Block or report deb888

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
deb888/README.md

AI Developer Banner

Typing Animation


🧬 SYSTEM IDENTITY

name: "Bruce Deb"
handle: "deb888"
role: "AI Developer · Fullstack Architect · AI DevOps Engineer"
mission: "Building AI-powered applications end-to-end — from React Native to LangChain agents deployed on Kubernetes"
philosophy: "Fullstack is the foundation. LangChain is the brain. Kubernetes is the body. OpenClaw is the heartbeat."

domains:
  mobile:       "React Native · Expo · Ionic/Capacitor — cross-platform native apps"
  ai_agents:    "LangChain · LangGraph · Deep Agents · Multi-Agent Systems · RAG · Tool-Use · Memory"
  agent_proto:  "MCP (Model Context Protocol) · A2A (Agent-to-Agent) · ACP (Agent Client Protocol)"
  frameworks:   "LangChain · CrewAI · Google ADK · Microsoft Agent Framework · Mastra · OpenAI Agents SDK"
  backend:      "NestJS · Fastify · Python · PostgreSQL · Redis · Kafka"
  frontend:     "Angular 18+ (Signals, Standalone) · React · Tailwind"
  aws:          "Bedrock · Lambda · API Gateway · DynamoDB · SQS · S3 — serverless AI on AWS"
  infra:        "Kubernetes · Terraform · ArgoCD · Istio · Crossplane"

tools_i_love:
  - "LangChain — The standard for LLM application development"
  - "LangGraph — Agent orchestration with cycles, state, and persistence"
  - "MCP — Model Context Protocol, the universal tool connector for AI"
  - "A2A — Google-led Agent-to-Agent protocol for multi-agent interoperability"
  - "Google ADK — Opinionated, batteries-included agent runtime for GCP"
  - "Microsoft Agent Framework — Unified AutoGen + Semantic Kernel at GA"
  - "Mastra — TypeScript-native agent framework with workflows + memory"
  - "Deep Agents — LangChain's long-running, persistent agent workflows"
  - "OpenAI Agents SDK — Clean multi-agent delegation with minimal abstraction"
  - "Claude Agent SDK — Anthropic's agent framework for tool-use and reasoning"
  - "OpenCode — AI coding agent I'm using right now to build this profile"
  - "OpenClaw — Personal AI assistant ecosystem (core contributor)"
  - "Gemini 2.5 / Claude 4 — The frontier models driving agent reasoning"
  - "React Native — One codebase, every platform"
  - "AWS Bedrock — Foundation models + agents as a managed service"
  - "AWS Serverless — Lambda, API Gateway, DynamoDB — pay-per-use AI backends"

ai_devops:
  ml_pipelines: "Kubeflow · MLflow · DVC · Weights & Biases"
  serving:      "AWS Bedrock · vLLM · TGI · Ollama · Triton Inference Server · TensorRT-LLM"
  monitoring:   "LangSmith · Prometheus · Grafana · Phoenix · Datadog"
  governance:   "Agent lifecycle management · Guardrails · Audit trails · Cost tracking"
  automation:   "OpenClaw Workflows · Cron · Heartbeat · Task Flow · Hooks"

⚡ THE AI ENGINEER STACK

graph TB
    subgraph "MOBILE FIRST"
        A["React Native / Expo"] --> B["iOS & Android"]
        A --> C["Hybrid (Ionic)"]
    end
    subgraph "AGENT PROTOCOLS"
        P1["🔌 MCP (Tool Access)"] --- P2["🤝 A2A (Agent-to-Agent)"]
        P2 --- P3["🔄 ACP (Agent Client)"]
    end
    subgraph "AI FRAMEWORKS 2026"
        D["LangChain"] --> E["LangGraph / Deep Agents"]
        F["Google ADK"] --> P2
        G["MS Agent Framework"] --> P1
        H["Mastra (TypeScript)"]
        I["OpenAI Agents SDK"]
        J["Claude Agent SDK"]
        K["CrewAI"]
        Z["☁️ AWS Bedrock"]
    end
    subgraph "MULTI-AGENT SYSTEMS"
        L["Supervisor / Worker"] --> M["Swarm / Pipeline"]
        M --> N["Reflection / ReAct"]
        N --> O["RAG · Tool-Use · Memory · Planning"]
    end
    subgraph "BACKEND & OPS"
        Q["NestJS / FastAPI"] --> R["PostgreSQL · Redis · Kafka"]
        Q --> S["REST · GraphQL · gRPC"]
        W["☁️ AWS Serverless"] --> Q
    end
    subgraph "AI DEVOPS & SERVING"
        T["Terraform"] --> U["Kubernetes"]
        U --> V["vLLM · Ollama · TGI · TensorRT-LLM"]
        V --> W["MLflow · Prometheus · LangSmith · Datadog"]
        W --> X["OpenClaw Workflow Engine"]
    end
    C --> Q
    O --> A & C & Q
    X --> U & L
    P1 --> D & K
    P3 --> X
    style A fill:#61DAFB,color:#000
    style D fill:#FFBE0B,color:#000
    style F fill:#4285F4,color:#fff
    style G fill:#00A4EF,color:#fff
    style H fill:#8B5CF6,color:#fff
    style O fill:#00FFE7,color:#000
    style P1 fill:#00C853,color:#fff
    style P2 fill:#FF6F00,color:#fff
    style T fill:#7B42BC,color:#fff
    style U fill:#326CE5,color:#fff
    style X fill:#FF006E,color:#fff
    style Z fill:#FF9900,color:#fff
    style W fill:#FF9900,color:#fff
Loading

🚀 PROJECTS THAT ACTUALLY DO THINGS

🦞 OpenClaw — Personal AI Assistant [Core Contributor]

382k+ ⭐ GitHub Self-hosted AI gateway connecting Telegram, WhatsApp, Discord, iMessage → AI agents

Stack:     TypeScript · Node.js · Plugin Architecture · MCP/ACP
My part:   Skills engine, automation layer, plugin SDK, multi-agent routing
Workflows: Cron jobs · Heartbeat monitoring · Task Flow orchestration · Hooks · Standing Orders

🦜️ LangChain AI Agent Suite [Building]

Multi-agent system with planning, tool-use, persistent memory, and human-in-the-loop

Stack:     LangChain · LangGraph · LangSmith · Qdrant · PostgreSQL
Patterns:  Supervisor/worker · Swarm · Pipeline · Reflection · ReAct

🔧 OpenCode + OpenClaw Integration [Workflow]

Using OpenCode (AI coding agent) paired with OpenClaw for autonomous development pipelines

Setup:  OpenCode writes code → OpenClaw deploys it → Cron monitors it → Heartbeat reports back
Flow:   "write a Slack bot" → OpenCode codes it → OpenClaw deploys to K8s → Cron keeps it alive

📱 React Native Mobile Apps

Cross-platform native apps — one TypeScript codebase, iOS + Android

Stack:    React Native · Expo · TypeScript · Socket.io
Projects: Real-time chat · GPS tracking · Image upload · REST APIs

☁️ AI DevOps Pipeline [Enterprise]

ML infrastructure: model serving, monitoring, auto-scaling, and cost optimization

Stack:     Kubernetes · Terraform · vLLM · MLflow · Prometheus · Grafana
Scale:     50+ clusters · 500+ services · 99.99% uptime

🔥 TRENDING NOW — 2026 AI Agent Stack

%%{init: {"theme": "base", "themeVariables": { "background": "#0D1117", "primaryColor": "#00FFE7", "primaryTextColor": "#fff", "primaryBorderColor": "#333", "lineColor": "#00FFE7", "secondaryColor": "#0a0a1a", "tertiaryColor": "#1a1a2e"}}}%%
graph LR
    subgraph "PROTOCOLS"
        A["🔌 MCP"] --> B["Tool Access"]
        C["🤝 A2A"] --> D["Agent Interop"]
        E["🔄 ACP"] --> F["Client Protocol"]
    end
    subgraph "FRAMEWORKS"
        G["LangGraph / Deep Agents"]
        H["Google ADK"]
        I["MS Agent Framework"]
        J["Mastra"]
        K["OpenAI Agents SDK"]
        L["Claude Agent SDK"]
        U["☁️ AWS Bedrock"]
    end
    subgraph "FRONTIER MODELS"
        M["⚡ Gemini 2.5 Pro"]
        N["🔥 Claude 4 Sonnet/Opus"]
        O["🧠 GPT-5"]
        P["🦙 Llama 4"]
    end
    subgraph "PRODUCTION"
        Q["vLLM · Ollama · TGI"]
        R["LangSmith · Datadog · Phoenix"]
        S["Kubernetes · Terraform · ArgoCD"]
        T["OpenClaw · Cron · Heartbeat"]
    end
    A & C & E --> G & H & I & J & K & L & U
    G & H & I & J & K & L & U --> M & N & O & P
    M & N & O & P --> Q
    Q --> R
    R --> S
    S --> T
    style A fill:#00C853,color:#fff
    style C fill:#FF6F00,color:#fff
    style E fill:#8B5CF6,color:#fff
    style G fill:#FFBE0B,color:#000
    style H fill:#4285F4,color:#fff
    style I fill:#00A4EF,color:#fff
    style M fill:#4285F4,color:#fff
    style N fill:#D97706,color:#fff
    style T fill:#FF006E,color:#fff
    style U fill:#FF9900,color:#fff
Loading

2026 Key Trends:

  • 🔌 MCP — Universal standard for AI-to-tool connectivity (Anthropic-led)
  • 🤝 A2A — Google-led Agent-to-Agent protocol for cross-framework interop
  • 🪟 Microsoft Agent Framework — AutoGen + Semantic Kernel unified at GA
  • 🎯 Google ADK — Batteries-included agent runtime for GCP-native stacks
  • 📦 Mastra — TypeScript-native agent framework (workflows, memory, Studio)
  • 🌀 Deep Agents — LangChain's long-running persistent agent workflows
  • 🔥 Claude 4 / Gemini 2.5 / GPT-5 — Frontier models driving agent reasoning
  • 🛡️ Agent Governance — Lifecycle management, guardrails, cost tracking

📊 COMPETENCY MATRIX

Competency Matrix


🧰 TOOLBOX

🦜️ LangChain & AI Agents (click to expand)
Tech What I Build With It
LangChain LLM app framework — chains, agents, RAG, tool-use, memory
LangGraph Stateful agent orchestration — cycles, branching, persistence
Deep Agents Long-running persistent agent workflows with LangChain
LangSmith LLM observability — traces, evaluations, prompt management
Google ADK Batteries-included agent runtime for GCP with A2A interop
MS Agent Framework Unified AutoGen + Semantic Kernel (GA) for .NET/enterprise
Mastra TypeScript-native agents with workflows, memory, and Studio
OpenAI Agents SDK Clean multi-agent delegation with minimal abstraction
Claude Agent SDK Anthropic's agent SDK — tool-use, reasoning, handoffs
CrewAI Role-based multi-agent teams — rapid prototyping
MCP Servers Model Context Protocol — universal tool connectivity
A2A Protocol Agent-to-Agent interoperability across frameworks
OpenClaw Personal AI assistants — skills engine, automation, plugin SDK
Qdrant / pgvector Vector stores for RAG — hybrid search, filtering, re-ranking
vLLM / Ollama / TGI Self-hosted model serving — privacy-first, local inference
AWS Bedrock Managed FM access — Claude, Llama, Mistral + agents via AWS
⚛️ React Native & Mobile (click to expand)
Layer Stack
Framework React Native, Expo, Ionic/Capacitor
Navigation React Navigation, Expo Router
State React Query, Zustand, Redux Toolkit
UI NativeBase, Tamagui, Tailwind Native
Native APIs Camera, Geolocation, Bluetooth, Push Notifications
Build EAS Build, Xcode, Android Studio, Fastlane
Store App Store Connect, Google Play Console
🔧 OpenCode & AI Development (click to expand)
Tech What I Do With It
OpenCode AI coding agent — building this profile and more
Claude Code / Codex Terminal-based AI coding assistants
OpenClaw + OpenCode Autonomous dev loop — OpenCode writes, OpenClaw deploys
MCP Servers Model Context Protocol — connecting AI to tools and data
A2A Protocol Agent-to-Agent — multi-framework agent coordination
ACP Agents Agent Client Protocol — multi-agent coordination
Genkit Google's open-source AI framework for TypeScript/Node.js
☸️ AI DevOps & MLOps (click to expand)
Tech What I Do With It
Kubernetes Model serving infrastructure — EKS/GKE/AKS
AWS Serverless Lambda · API Gateway · DynamoDB · SQS · S3 — serverless AI backends
vLLM / TGI / TensorRT-LLM LLM inference serving with continuous batching
Terraform / Pulumi / Crossplane Infrastructure as Code for ML platforms
MLflow / W&B Experiment tracking, model registry, artifact storage
Prometheus / Grafana / Datadog Model monitoring — latency, throughput, token usage
LangSmith / Phoenix LLM observability — traces, evals, prompt management
ArgoCD / Flux GitOps for model deployments
Kubeflow / Flyte ML pipelines on Kubernetes
Agent Governance Lifecycle mgmt, guardrails, audit trails, cost tracking
🦞 OpenClaw Workflow Engine (click to expand)
Feature What It Does
Cron Jobs Scheduled tasks — daily reports, reminders, batch jobs
Heartbeat Periodic agent turns every ~30 min — inbox check, calendar scan
Task Flow Durable multi-step orchestration with revision tracking
Hooks Event-driven scripts — session lifecycle, message flow
Standing Orders Persistent agent instructions (AGENTS.md)
Inferred Commitments Memory-like follow-ups from natural conversation
Skills SKILL.md instruction packs — teach agents new abilities
Plugins Channel, tool, and provider extensions from ClawHub
⚛️ Full Stack (click to expand)
Layer Stack
Frontend Angular 18+ (Signals, Standalone), React, Tailwind
Backend NestJS, Fastify, Express, Python (FastAPI), Go
Database PostgreSQL, MongoDB, Redis, TimescaleDB, ClickHouse
API REST, GraphQL Federation, gRPC, tRPC, WebSocket
Mobile React Native, Expo, Ionic/Capacitor
Testing Vitest, Playwright, Cypress, Testing Library

📊 ACTIVITY

Stats Languages

Streak

Trophies


🎯 NOW

mindmap
  root((🧠 Current Focus))
    Agent Protocols
      MCP
      A2A
      ACP
    AI Frameworks 2026
      LangGraph
      Deep Agents
      Google ADK
      MS Agent Framework
      Mastra
      AWS Bedrock
      OpenAI Agents SDK
      Claude Agent SDK
    Frontier Models
      Gemini 2.5 Pro
      Claude 4 Sonnet
      GPT-5
      Llama 4
    React Native
      Cross-platform apps
      Expo
      Native modules
    OpenClaw
      Skills
      Workflows
      Plugin SDK
    OpenCode
      AI coding
      Agent protocols
      MCP
    AI DevOps
      K8s model serving
      AWS Serverless
      ML monitoring
      Agent governance
Loading

📡 CONNECT

AI Developer · Fullstack · AI DevOps · OpenCode · OpenClaw
"React Native → LangChain → Kubernetes → OpenClaw workflow — I build the whole pipeline."


Popular repositories Loading

  1. node-muter-imageupload-rest-api node-muter-imageupload-rest-api Public

    TypeScript 1

  2. node-chat-socket.io-ionic node-chat-socket.io-ionic Public

    JavaScript 1

  3. ionic-chat-app ionic-chat-app Public

    JavaScript 1

  4. trackinionic trackinionic Public

    JavaScript 1

  5. custom-npm-test custom-npm-test Public

  6. bacupcode bacupcode Public

    all backup

    TypeScript