Commit History
-
feat(phase1): rebuild backend on FastAPI + E2B + SSE (stability-first)
genspark-ai-developerComplete Phase-1 rebuild per spec: one stable pipeline, real execution, no fake Computer Use. Backend (backend/app.py — single ~530-line module): - GET /health: real status (E2B + provider availability) - POST /api/v1/chat (SSE): LLM-only streaming, NO sandbox * Providers in order: SambaNova → Groq → GitHub Models → OpenAI → Anthropic → Gemini * Stable text-only streaming, no fragile tool-calling - POST /api/v1/execute (SSE): REAL E2B sandbox execution * Streams sandbox_ready → stdout/stderr → result events live * Per-session sandbox reuse (warm restart ~88ms vs 400ms cold) * Accepts {language, code} OR legacy {tool, args} * Local subprocess fallback if E2B unavailable - POST /api/v1/agent (SSE): intent-routed (chat OR execute) * Keyword + code-block heuristic, LLM code generation when needed - POST /api/v1/orchestrate, /api/v1/kernel/orchestrate, /api/v1/chat/stream: legacy aliases for frontend backward-compat - GET/DELETE /api/v1/sandbox/{session_id}: lifecycle - WS …bb9b14c
-
feat: God Agent OS v9 — Space-Role Architecture
Pyae Sone- Agent Kernel: replaces GodAgentOrchestratorV7 with generalized kernel - 8 Spaces: Core, Browser, Sandbox, Coding, Vision, Debug, Deploy, Communication - 5 Roles: Cognition, Automation, Execution, Repair, Visual Intelligence - New frontend: Space-Role UI with connectors panel, spaces visualization - Dashboard: real-time Space activation, role indicator, chat interface - TopBar: 'Powered by Pyae Sone' branding - HuggingFace Dockerfile.hf + README.md for HF Space deployment - Vercel: vercel.json with API proxy to HF backend - GitHub Actions: auto-deploy workflow - Backward compatible with v8 16-agent fleet Powered by Pyae Sone11b240f -
feat: God Agent OS v7 — Autonomous Engineering OS (Manus+Genspark+Devin)
God Agent v7🤖 NEW v7 Agents (5 new): - BrowserAgent: Web research & scraping via DuckDuckGo + httpx - FileAgent: Full file system control, project scaffolding, tree view - GitAgent: Autonomous Git ops, GitHub PR creation, code review - TestAgent: Auto pytest/Jest test generation & execution - VisionAgent: Design-to-code UI generation (dashboard, landing, components) ⚡ Core Upgrades: - orchestrator_v7.py: 16-agent parallel execution, advanced intent classification - main_v7.py: New entry point registering all 16 agents - Self-healing loop with asyncio.wait_for timeouts - Enhanced WebSocket streaming for all new agents 🎨 Frontend v7: - FileExplorer panel: Live workspace file tree with icons - BrowserPanel: Web research UI with history & quick searches - Updated Sidebar: all 16 agents shown with v7 badges - Updated ChatPanel: v7 quick actions (Research, Scaffold, Git, Test, Vision) - useAgentStore: Extended ActivePanel types (files, browser) 🚀 DevOps & Deploy: - .github/workflows/deploy.yml: CI…5d62489
-
feat: GOD MODE+ v3.0 — Full Autonomous AI Operating System
pyaesonegtckglay-dotcomPHASE 1: God Agent Orchestrator - GodAgentOrchestrator: central brain with intent classification - 10 specialized agents: Chat, Planner, Coding, Debug, Memory, Connector, Deploy, Workflow, Sandbox, UI - Task delegation engine with parallel agent execution - Self-healing retry loop (DebugAgent) PHASE 2: Persistent VS Code Sandbox - SandboxAgent: file I/O, terminal exec, git operations - /ws/sandbox/{session_id} WebSocket live terminal - Sandbox REST API: /api/v1/agents/sandbox/* - Workspace persistence at /tmp/god_workspace PHASE 3: Connector System (13 connectors) - ConnectorManager: GitHub, HF, Vercel, OpenAI, Groq, Cerebras, OpenRouter, Anthropic, n8n, Telegram, Discord, Slack, Cloudflare - ConnectorsPanel UI (Manus-style icon grid) - Runtime token management API PHASE 4: Autonomous Coding Engine - CodingAgent: multi-file generation, refactoring, repo scan - Self-healing loop: generate→test→detect→retry→fix - DebugAgent: deep error analysis, syntax validation PHASE 5: Memor…a20767f
-
🚀 Devin Agent Platform v2.0
pyaesonegtckglay-dotcom666aab6