GeekSpace2.0

<div align="center">

<img src="docs/assets/banner.svg" alt="GeekSpace 2.0" width="100%" />

<br />
<br />

[![Live](https://img.shields.io/badge/LIVE-ai.agentin.chat-7B61FF?style=for-the-badge&logo=data:image/svg+xml;base64,PHN2ZyB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciIHdpZHRoPSIxNiIgaGVpZ2h0PSIxNiIgdmlld0JveD0iMCAwIDI0IDI0IiBmaWxsPSJub25lIiBzdHJva2U9IiNmZmYiIHN0cm9rZS13aWR0aD0iMiI+PHBhdGggZD0iTTEyIDJMMTUuMDkgOC4yNiAyMiA5LjI3IDE3IDEzLjE0IDE4LjE4IDIxLjAyIDEyIDE3LjI3IDUuODIgMjEuMDIgNyAxMy4xNCAyIDkuMjcgOC45MSA4LjI2IDEyIDJaIi8+PC9zdmc+)](https://ai.agentin.chat)
[![Version](https://img.shields.io/badge/v3.3.0-platform-61FF7B?style=for-the-badge)](https://github.com/trendywink247-afk/GeekSpace2.0/releases)
[![CI](https://img.shields.io/github/actions/workflow/status/trendywink247-afk/GeekSpace2.0/ci.yml?branch=main&style=for-the-badge&label=CI&logo=github)](https://github.com/trendywink247-afk/GeekSpace2.0/actions)
[![Status](https://img.shields.io/badge/Status-status.agentin.chat-61FF7B?style=for-the-badge)](https://status.agentin.chat)
[![License](https://img.shields.io/badge/License-MIT-61FF7B?style=for-the-badge)](LICENSE)

[![React](https://img.shields.io/badge/React_19-61DAFB?style=flat-square&logo=react&logoColor=black)](https://react.dev)
[![TypeScript](https://img.shields.io/badge/TypeScript-3178C6?style=flat-square&logo=typescript&logoColor=white)](https://www.typescriptlang.org/)
[![Express](https://img.shields.io/badge/Express-000000?style=flat-square&logo=express&logoColor=white)](https://expressjs.com)
[![SQLite](https://img.shields.io/badge/SQLite-003B57?style=flat-square&logo=sqlite&logoColor=white)](https://www.sqlite.org/)
[![Docker](https://img.shields.io/badge/Docker-2496ED?style=flat-square&logo=docker&logoColor=white)](https://www.docker.com/)
[![Tailwind](https://img.shields.io/badge/Tailwind-06B6D4?style=flat-square&logo=tailwindcss&logoColor=white)](https://tailwindcss.com/)
[![Playwright](https://img.shields.io/badge/Playwright-2EAD33?style=flat-square&logo=playwright&logoColor=white)](https://playwright.dev)
[![Node 20](https://img.shields.io/badge/Node_20-339933?style=flat-square&logo=nodedotjs&logoColor=white)](https://nodejs.org/)
[![PWA](https://img.shields.io/badge/PWA-5A0FC8?style=flat-square&logo=pwa&logoColor=white)](#)
[![Ollama](https://img.shields.io/badge/Ollama-000000?style=flat-square&logo=ollama&logoColor=white)](https://ollama.ai)

**A self-hosted AI OS — your agent, your dashboard, your portfolio.**

> v3.3.0 · Agentic Experience · Goals + Delegation + Deep Reasoning · 2552 tests · Security-hardened · 15+ Docker services · PR-based CI/CD

[Live Demo](https://ai.agentin.chat) · [Documentation](docs/) · [Report Bug](.github/ISSUE_TEMPLATE/bug_report.yml) · [Request Feature](.github/ISSUE_TEMPLATE/feature_request.yml)

</div>

---

## What Is Agentin?

Agentin is a personal AI platform that gives every user their own intelligent agent, public developer portfolio, built-in terminal, and automation engine. Route queries across local and cloud AI, swap personalities, automate workflows, and share your work — all from a single self-hosted dashboard.

**No vendor lock-in. No data leaving your infra unless you choose it.**

### What makes it different

- **Local-first AI** — Ollama runs on your hardware. Cloud is optional fallback, not the default.
- **One platform, not 10 tools** — Agent, portfolio, reminders, automations, billing, terminal — unified.
- **9 AI personalities** — Weebo, Edith, Jarvis, Aria, Forge, Pulse, Echo, Cal, Nova — switch mid-message.
- **Auto-delegation** — Weebo detects intent and routes to the right specialist agent automatically.
- **Multi-agent council** — Fan-out to parallel specialists for complex queries.
- **Goal-driven autonomy** — Set goals, agents decompose into steps, pursue them autonomously, report progress.
- **Inter-agent delegation** — Agents hand off work to specialists mid-conversation with full audit trail.
- **Deep reasoning** — Complex queries get 10-iteration reasoning with self-reflection and plan-then-execute.
- **Proactive goal engine** — Agents auto-execute steps, nudge stale goals, send daily progress summaries.
- **Telegram-native** — Full AI on Telegram: voice notes, inline keyboards, receipt OCR, proactive nudges.
- **Hinglish-first** — Built for Indian users. "swiggy pe 350 rupay" just works.
- **Background agents** — Weebo Engine + proactive briefings, habit nudges, expense digests.
- **Credit economy** — Fair usage with transparent per-call costs and multiple pricing tiers.
- **Security-hardened** — 28 findings audited and fixed (Session 9). JWT 15m expiry, no hardcoded secrets, `no-new-privileges` on all containers.

---

## Features

<table>
<tr>
<td width="50%">

**AI Agent & Chat**
- 9 Personalities — Weebo, Edith, Jarvis, Aria, Forge, Pulse, Echo, Cal, Nova
- Named agent routing: "hey Aria", "@Nova", "Forge:" switches mid-message
- Cloud-first LLM waterfall: OpenRouter-free → PicoClaw → Ollama → Groq → Together → Maverick → Kimi → Edith
- PicoClaw circuit breaker (1 failure → 5min cooldown)
- Auto-delegation: Weebo routes to Cal/Echo/Forge/Aria/Pulse/Nova/Jarvis by intent
- Multi-Agent Council — "launch mode" fan-out to 3 parallel specialists
- ReAct loop with 23 tools (notes, habits, reminders, expenses, focus, briefings, goals, artifacts, etc.)
- Deep Reasoning Engine — 10 iterations, self-reflection, plan-then-execute for complex queries
- Goal System — create goals, AI decomposes into steps, agents pursue autonomously
- Inter-Agent Delegation — agents hand off work to specialists with audit trail
- Workspace Artifacts — shared scratchpad for inter-agent collaboration
- Hinglish routing — Indian language patterns + merchant auto-categories
- Long-term memory — per-user fact store, auto-injected into prompts
- Conversation context preserved across long replies (16K char window)
- SSE streaming responses
- Chat search, export, reactions, AI model preference

</td>
<td width="50%">

**Developer Portfolio**
- Public profile at `username.agentin.chat`
- AI-powered visitor chat (talk to someone's agent)
- Project showcase with AI-generated descriptions
- Connection tracking and social links
- Multiple layout themes
- Portfolio public sharing with visit analytics

</td>
</tr>
<tr>
<td>

**Automations & Agentic Engine**
- Weebo Engine — up to 3 background agents per user
- Proactive Goal Engine — auto-executes goal steps, nudges stale goals, daily summaries
- Agent-Initiated Notifications — milestone alerts at 25/50/75/100% via SSE + Telegram
- Autonomy Levels — manual, suggest, semi_auto, full_auto (user controls how proactive agents are)
- Proactive Engine V3 — 30-min reminder previews, habit idle nudges at 11:00 IST
- Daily briefings with habit insights + active streaks + at-risk habits
- Telegram integration — inline keyboards (Done/Snooze/Delete), photo vision, file/doc handling
- Reminders via push, email, or Telegram; smart recurrence (daily/weekly/monthly)
- Expense tracker — track spend, categories, budget limits, weekly digest
- Habit Intelligence V2 — streak tracking, status icons, motivational nudges
- Focus/Pomodoro sessions, note-taking, global search across all data
- Cron, webhook, and health-check triggers

</td>
<td>

**Dashboard & Tools**
- Real-time stats, credit usage, activity feed
- Built-in terminal with `gs` commands and `ai "..."` queries
- API key management with AES-256-GCM encryption
- Billing with INR/USD pricing (Free → Yearly plans)
- PWA — installable on mobile and desktop
- Google OAuth — Calendar + Gmail + Sign-in
- Focus mode, habits tracker, personal analytics
- Auth session management (view and revoke active sessions)
- Activity notification log

</td>
</tr>
<tr>
<td colspan="2">

**Office Page**
- Pixel-perfect agent sprites (16x32, 3-row layout, integer 2x scaling)
- BFS pathfinding with collision map, room zones, smart objects
- Activity stream with 12 event types + agent attribution
- 60/40 layout: canvas left, S

[truncated…]