v1.2.0: opencode JSON mode, 180s Hermes timeout, client AbortController, SEO README, badge/CSS fixes, prompt clipboard fix

2026-05-17 23:43:18 +05:30 · 2026-05-17 23:43:18 +05:30 · 2b3083308f
parent 828a6ab38a
commit 2b3083308f
14 changed files with 489 additions and 186 deletions
--- a/.gitignore
+++ b/.gitignore
@ -7,3 +7,7 @@ node_modules/
 .DS_Store
 *.tar.gz
 backups/*.tar.gz
+audit/*
+data/settings.json
+data/chat-history.json
+data/cost-history.json
--- a/AGENTS.md
+++ b/AGENTS.md
@ -14,6 +14,7 @@ Your role is to act as the **kernel** of this system: route tasks to the right a
 |-------|-------|
 | **Name** | Agentic OS |
 | **Location** | `/home/mihir/Desktop/Agentic OS Project/` |
+| **GitHub** | [github.com/modimihir07/agentic-os](https://github.com/modimihir07/agentic-os) |
 | **Author** | Mihir N Modi (Gujarat, India — BTech CSE 1st year) |
 | **Created** | May 17, 2026 |
 | **License** | MIT |
@ -314,7 +315,7 @@ Your role is to act as the **kernel** of this system: route tasks to the right a
 | **Active Project** | CloudMart — GCP DevOps multi-region e-commerce platform |
 | **Other Projects** | AgriAssist AI, Hermes Agent/OpenClaw, EROS Wellness AI, Java OOP practice |
 | **CLI Tools Available** | opencode, Hermes Agent, Gemini CLI |
-| **Preferred Model** | deepseek-v4-flash-free |
+| **Preferred Model** | Hermes: Owl Alpha (OpenRouter, free), opencode: deepseek-v4-flash-free (opencode-zen), Gemini: gemini-2.5-flash (Google OAuth) |
 | **Obsidian Vaults** | 4 vaults: DevOps, DeepSeek Chat, AI/ML, SEM-2 Academics (192 files) |

 ---
@ -343,6 +344,11 @@ When you (an AI agent) are dropped into this directory for the first time:
 - Any AI agent reading this file should have enough context to continue seamlessly
 - The agent should first read this file, then `brain/*`, then `data/settings.json`

+### Session History
+- opencode: `~/.local/share/opencode/opencode.db` + `~/.local/share/opencode/log/`
+- Hermes: `~/.hermes/sessions.json`
+- Gemini: `~/.gemini/history/`
+
 ---

 ## API Endpoints (server.py FastAPI)
@ -379,6 +385,7 @@ When you (an AI agent) are dropped into this directory for the first time:
 | Date | Version | Changes |
 |------|---------|---------|
 | May 17, 2026 | v1.0.0 | Initial creation — all 51 features + 10 extras |
+| May 17, 2026 | v1.1.0 | AI Chat fixes: corrected CLI syntax (hermes `chat -q`, gemini positional), switched Hermes default to OpenRouter + Owl Alpha, improved timeout/error handling, updated execute_agent fallback logic |

 ---

--- a/README.md
+++ b/README.md
@ -1,42 +1,283 @@
-# Agentic OS
+<div align="center">
+  <br/>
+  <img src="https://img.shields.io/badge/license-MIT-blue.svg" alt="License: MIT"/>
+  <img src="https://img.shields.io/badge/python-3.10+-blue.svg" alt="Python 3.10+"/>
+  <img src="https://img.shields.io/badge/FastAPI-0.115+-green.svg" alt="FastAPI"/>
+  <img src="https://img.shields.io/badge/agents-3-orange.svg" alt="3 Agents"/>
+  <img src="https://img.shields.io/badge/skills-16-purple.svg" alt="16 Skills"/>
+  <img src="https://img.shields.io/badge/status-stable-brightgreen.svg" alt="Status: Stable"/>
+  <br/><br/>
+</div>

-A multi-agent orchestration platform that coordinates **opencode**, **Hermes Agent**, and **Gemini CLI** into a unified, self-improving, autonomous work operating system.
+# Agentic OS 🧠 — Multi-Agent Orchestration Platform

-## Quick Start
+A locally-hosted operating system for AI agents that coordinates **opencode**, **Hermes Agent**, and **Gemini CLI** into a unified dashboard with persistent memory, cron scheduling, skill execution, cost analytics, and disaster recovery.
+
+> **Why Agentic OS?** Most agent tools work in isolation — a terminal for coding, a separate chat for research, another for memory. Agentic OS is the **control plane** that unifies them: one dashboard, one memory layer, one scheduler, one skill hub. Three agents, infinite capabilities.
+
+---
+
+## ✨ Features
+
+| Category | Features |
+|----------|----------|
+| **🤖 3-Agent Engine** | opencode (code/DevOps), Hermes (memory/scheduling), Gemini (research/analysis) with intelligent routing |
+| **🧩 16+ Skills** | Executable skill packs with eval scoring, learnings, and score history per run |
+| **🧠 Persistent Memory** | SQLite FTS5 + `brain/` folder — shared context read by all agents at session start |
+| **⏱ Cron Scheduler** | APScheduler-powered jobs — heartbeat, memory consolidation, daily standup, DevOps audit |
+| **💰 Cost Analytics** | Track spending per provider, model, agent. Free-tier alerts prevent surprise bills |
+| **📋 Audit Trail** | Every action logged — chat, skill runs, settings changes, backups |
+| **💾 One-Click Backup** | Full tar.gz snapshot of all configs, skills, brain, agents, prompts |
+| **📝 Prompt Library** | 10 reusable templates — code review, system audit, project plan, brainstorm, and more |
+| **📐 Standards System** | Discover and inject coding conventions across your project |
+| **🔌 Plugin Registry** | Marketplace-style plugin management (extensible via skills) |
+| **🎨 Dark/Light Theme** | GitHub-style dark mode + clean light theme, toggle from sidebar |
+| **⚡ Zero API Costs** | Built for free tiers — Gemini Flash, OpenRouter free models, local opencode |
+
+---
+
+## 🏗 Architecture
+
+```
+┌──────────────────────────────────────────────────────────────┐
+│                    AGENTIC OS DASHBOARD                       │
+│                    FastAPI + Tailwind SPA                     │
+├──────────────────────────────────────────────────────────────┤
+│                                                               │
+│  ┌──────────────┐  ┌──────────────┐  ┌──────────────────┐   │
+│  │   opencode    │  │    Hermes    │  │   Gemini CLI     │   │
+│  │  (Code/DevOps)│  │ (Memory/Sched│  │ (Research/Analy) │   │
+│  │  File Ops)    │  │  /Channels)  │  │                  │   │
+│  └──────────────┘  └──────────────┘  └──────────────────┘   │
+│                                                               │
+│  ┌──────────────────────────────────────────────────────┐    │
+│  │              7 CORE LAYERS (Stacked)                  │    │
+│  │  Layer 7: Identity / Persona / Constitution           │    │
+│  │  Layer 6: Self-Evolution + Capability Manager         │    │
+│  │  Layer 5: Scheduler + Awareness + Health Guardian     │    │
+│  │  Layer 4: Memory Graph + Memory Consolidation         │    │
+│  │  Layer 3: Skills Hub + Eval + Learnings Loop          │    │
+│  │  Layer 2: Business Brain + Context Folders            │    │
+│  │  Layer 1: Agent Router + Standards + Profiles         │    │
+│  └──────────────────────────────────────────────────────┘    │
+└──────────────────────────────────────────────────────────────┘
+```
+
+### Agent Responsibilities
+
+| Agent | Role | Default Model | Provider | Cost |
+|-------|------|---------------|----------|------|
+| **opencode** | Code generation, DevOps, file operations | deepseek-v4-flash-free | opencode-zen | **$0** |
+| **Hermes Agent** | Persistent memory, scheduling, messaging | Owl Alpha (1M ctx) | OpenRouter | **$0** |
+| **Gemini CLI** | Web research, multi-modal analysis | gemini-2.5-flash | Google OAuth | **$0** |
+
+### Routing Rules
+
+- **Code/DevOps task?** → opencode
+- **Memory/Channel/Schedule?** → Hermes Agent
+- **Research/Analysis?** → Gemini CLI
+- **Complex multi-step?** → Chain: Gemini researches → opencode implements → Hermes monitors/schedules
+
+---
+
+## 🚀 Quick Start

 ```bash
+git clone https://github.com/modimihir07/agentic-os.git
+cd agentic-os
 chmod +x install.sh && ./install.sh
-```
-
-## Architecture
-
-3 agents + 7 layers + web dashboard:
-
-| Agent | Role |
-|-------|------|
-| **opencode** | Code generation, DevOps, file operations |
-| **Hermes Agent** | Memory, scheduling, channels |
-| **Gemini CLI** | Research, analysis, multi-modal |
-
-## Features
-
- 51 features from top agentic OS implementations
- 10 extra features (cost analytics, git versioning, backup, prompts, etc.)
- Interactive web dashboard (FastAPI + SPA)
- One-click skill execution
- Scheduled workflows
- Persistent memory across sessions
- Self-improving skills with eval scoring
- Multi-agent routing and handoff
- Cost tracking with free-tier alerts
-
-## Dashboard
-
-```bash
 ./start.sh
-# Visit http://127.0.0.1:8080
+# Open http://127.0.0.1:8080
 ```

-## License
+---

-MIT
+## 📋 Prerequisites
+
+| Tool | Required? | Install |
+|------|-----------|---------|
+| Python 3.10+ | ✅ Required | `sudo apt install python3 python3-pip` |
+| Node.js 18+ | ⚠ For opencode | `curl -fsSL https://deb.nodesource.com/setup_20.x \| sudo bash - && sudo apt install -y nodejs` |
+| opencode | ⚠ For code tasks | `npm install -g @opencode/cli` |
+| Hermes Agent | ⚠ For memory/scheduling | `curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh \| bash` |
+| Gemini CLI | ⚠ For Google AI | `npm install -g @google/gemini-cli` |
+
+> ⚠ = Optional — the dashboard works with any subset of installed agents.
+
+---
+
+## 🔧 Configuration
+
+### Hermes (OpenRouter)
+```bash
+# Set your OpenRouter API key
+echo 'OPENROUTER_API_KEY=sk-or-v1-your-key-here' > ~/.hermes/.env
+```
+
+### Gemini CLI (Google OAuth)
+```bash
+gemini auth login
+# Complete OAuth in browser — tokens saved to ~/.gemini/oauth_creds.json
+```
+
+### Dashboard Settings
+Edit `data/settings.json`:
+```json
+{
+  "dashboard": { "port": 8080 },
+  "theme": "dark",
+  "agents": { "opencode": true, "hermes": true, "gemini": true }
+}
+```
+
+---
+
+## 📁 Project Structure
+
+```
+agentic-os/
+├── server.py              # FastAPI backend (REST API)
+├── requirements.txt       # Python dependencies
+├── install.sh             # One-command installer
+├── start.sh               # Launch dashboard
+├── backup.sh / restore.sh # Disaster recovery
+│
+├── dashboard/             # Web frontend (SPA)
+│   ├── index.html         # Entry point + sidebar
+│   ├── app.js             # SPA router + theme/sidebar toggle
+│   ├── api.js             # API client (all endpoints)
+│   ├── styles.css         # Full dark/light theme CSS
+│   ├── utils.js           # Shared utilities
+│   └── pages/             # 13 page modules
+│       ├── dashboard.js   # Overview with stats
+│       ├── skills.js      # Skill grid/list/detail
+│       ├── memory.js      # Brain file editor
+│       ├── chat.js        # Multi-agent chat
+│       ├── scheduler.js   # Cron job manager
+│       ├── audit.js       # Activity trail
+│       ├── cost.js        # Cost analytics charts
+│       ├── plugins.js     # Plugin marketplace
+│       ├── backups.js     # Backup/restore UI
+│       ├── prompts.js     # Template library
+│       ├── standards.js   # Code conventions
+│       ├── settings.js    # Config editor
+│       └── setup-wizard.js  # Guided setup
+│
+├── brain/                 # Shared context (all agents read)
+│   ├── business-brain.md  # Current project context
+│   ├── memory.md          # Accumulated knowledge
+│   ├── recent-decisions.md
+│   ├── active-projects.md
+│   ├── identity.md
+│   └── constitution.md
+│
+├── skills/                # 16 executable skills
+│   ├── devops-audit/      # GCP/CloudMart infra audit
+│   ├── heartbeat/         # 5-min health check
+│   ├── content-draft/     # Blog/newsletter writing
+│   ├── code-review/       # PR review checklist
+│   ├── research-synthesis/ # Gemini research aggregator
+│   ├── daily-standup/     # Morning briefing
+│   ├── meeting-minutes/   # Notes processor
+│   ├── project-planner/   # Step-by-step plans
+│   ├── brainstorming/     # Socratic design
+│   ├── systematic-debug/  # 4-phase root cause
+│   ├── memory-consolidation/ # Weekly synthesis
+│   ├── backup-skill/      # Snapshot creator
+│   ├── cost-analytics/    # Multi-provider cost
+│   ├── tdd-cycle/         # Red-green-refactor
+│   ├── goal-planner/      # Goal → steps
+│   └── _template/         # Starter template
+│
+├── agents/                # Per-agent configs
+├── scheduler/jobs/        # Cron job definitions
+├── registry/              # Plugin marketplace
+├── standards/             # Discover/inject conventions
+├── prompts/               # 10 reusable templates
+├── data/                  # Runtime data (gitignored)
+├── audit/                 # Activity log (gitignored)
+└── backups/               # Snapshots (gitignored)
+```
+
+---
+
+## 🎮 Usage
+
+### AI Chat
+Select an agent from the sidebar → type your message → get response.
+
+| Agent | Best for | Example |
+|-------|----------|---------|
+| **opencode** | "Check system status", "Deploy to GKE" | Code + terminal automation |
+| **Hermes** | "What did I work on recently?", "Schedule a daily backup" | Memory recall, scheduling |
+| **Gemini** | "Research latest AI agent trends", "Analyze this image" | Web research, multi-modal |
+
+### Skills
+Browse 16 skills from Skills Hub → click Run → monitor eval scores over time.
+
+### Scheduler
+Create cron jobs: heartbeat (5 min), memory consolidation (weekly), daily standup, DevOps audit.
+
+### Cost Analytics
+Track spending per provider/model/agent. Free-tier alerts warn when nearing limits.
+
+---
+
+## 📊 Comparison: Agentic OS vs Claude Agent OS (Julian Goldie)
+
+| Feature | Claude Agent OS (Video) | Agentic OS (This Project) |
+|---------|------------------------|---------------------------|
+| **Core Agents** | Claude + OpenClaw + Hermes | opencode + Hermes + Gemini CLI |
+| **Cost** | $20/mo (Claude subscription) | **$0 — all free tiers** |
+| **Stack** | Next.js + Tailwind | FastAPI + vanilla JS SPA |
+| **Architecture** | 4 layers | **7 layers** |
+| **Skills System** | Plugin marketplace (2,000+ from Hermes) | 16 curated skills + eval scoring + learnings |
+| **Memory** | Obsidian vault (external) | Built-in brain/ + SQLite FTS5 |
+| **Scheduler** | Not shown | APScheduler cron jobs |
+| **Cost Tracking** | Not shown | Built-in per-provider analytics |
+| **Backup/Restore** | Not shown | One-click tar.gz snapshots |
+| **Audit Trail** | Not shown | Full activity log |
+| **Standards System** | Not shown | Discover/inject conventions |
+| **Client Timeout** | Not shown | 200s AbortController |
+| **Kanban Board** | Yes | No (not needed for agent OS) |
+| **Open Source** | No (tutorial only) | **MIT License** |
+
+---
+
+## 🧪 Tested On
+
+- **OS**: Linux (Ubuntu 22.04+), macOS
+- **Python**: 3.10, 3.11, 3.12
+- **Browsers**: Chrome, Firefox, Edge
+- **Agents**: opencode v0.8+, Hermes Agent v1.0+, Gemini CLI v1.0+
+
+---
+
+## 🤝 Contributing
+
+1. Fork the repo
+2. Create a feature branch (`git checkout -b feature/amazing-feature`)
+3. Commit your changes (`git commit -m 'Add amazing feature'`)
+4. Push (`git push origin feature/amazing-feature`)
+5. Open a Pull Request
+
+See `data/agent-routes.json` for routing rules and `skills/_template/` for creating new skills.
+
+---
+
+## 📄 License
+
+MIT License — free to use, modify, and distribute.
+
+---
+
+<div align="center">
+  <p><strong>Agentic OS</strong> — Built with ❤️ by <a href="https://github.com/modimihir07">Mihir Modi</a></p>
+  <p>
+    <a href="https://github.com/modimihir07/agentic-os/issues">Report Bug</a>
+    ·
+    <a href="https://github.com/modimihir07/agentic-os/discussions">Discussions</a>
+    ·
+    <a href="https://github.com/modimihir07/agentic-os">GitHub</a>
+  </p>
+</div>
--- a/audit/audit.log
+++ b/audit/audit.log
@ -1,8 +0,0 @@
-{"action": "skill_run", "skill": "heartbeat", "agent": "opencode", "run_id": "d52d517a", "timestamp": "2026-05-17T09:58:07.559900+00:00", "id": "36284cbc"}
-{"action": "skill_run", "skill": "devops-audit", "agent": "opencode", "run_id": "92cd4f97", "timestamp": "2026-05-17T09:59:08.213032+00:00", "id": "40695cd7"}
-{"action": "skill_run", "skill": "heartbeat", "agent": "opencode", "run_id": "f18c903b", "timestamp": "2026-05-17T10:18:44.356138+00:00", "id": "c9117c6d"}
-{"action": "job_deleted", "job_id": "daily-standup-default", "timestamp": "2026-05-17T10:34:38.978261+00:00", "id": "35513486"}
-{"action": "chat_message", "agent": "opencode", "msg_preview": "hello", "timestamp": "2026-05-17T10:37:06.791873+00:00", "id": "c41e3054"}
-{"action": "chat_message", "agent": "opencode", "msg_preview": "hello", "timestamp": "2026-05-17T10:40:56.501310+00:00", "id": "6d4ab33d"}
-{"action": "chat_message", "agent": "opencode", "msg_preview": "hello", "timestamp": "2026-05-17T10:42:36.570364+00:00", "id": "358f4058"}
-{"action": "chat_message", "agent": "opencode", "msg_preview": "hey", "timestamp": "2026-05-17T10:49:19.673600+00:00", "id": "c188ba46"}
--- a/dashboard/api.js
+++ b/dashboard/api.js
@ -4,8 +4,10 @@ const api = {
    if (!r.ok) { const e = await r.json().catch(() => ({})); throw new Error(e.detail || `Request failed: ${r.status}`); }
    return r.json();
  },
-  async post(path, body = {}) {
-    const r = await fetch(path, { method: 'POST', headers: { 'Content-Type': 'application/json' }, body: JSON.stringify(body) });
+  async post(path, body = {}, controller) {
+    const opts = { method: 'POST', headers: { 'Content-Type': 'application/json' }, body: JSON.stringify(body) };
+    if (controller) opts.signal = controller.signal;
+    const r = await fetch(path, opts);
    if (!r.ok) { const e = await r.json().catch(() => ({})); throw new Error(e.detail || `Request failed: ${r.status}`); }
    return r.json();
  },
@ -43,6 +45,6 @@ const api = {
  updateSettings: (settings) => api.put('/api/settings', { settings }),
  getStandards: () => api.get('/api/standards'),
  discoverStandards: () => api.post('/api/standards/discover'),
-  chat: (agent, message) => api.post('/api/chat', { agent, message }),
+  chat: (agent, message, controller) => api.post('/api/chat', { agent, message }, controller),
  getChatHistory: () => api.get('/api/chat/history'),
 };
--- a/dashboard/app.js
+++ b/dashboard/app.js
@ -28,6 +28,10 @@ async function navigate(page) {
  const hash = page || window.location.hash.slice(1) || 'dashboard';
  if (!hash) { window.location.hash = 'dashboard'; return; }

+  // Show loading bar
+  const bar = document.getElementById('topLoadingBar');
+  if (bar) { bar.classList.add('active'); bar.style.width = '30%'; }
+
  document.querySelectorAll('.nav-item').forEach(el => el.classList.remove('active'));
  const navItem = document.querySelector(`[data-page="${hash}"]`);
  if (navItem) navItem.classList.add('active');
@ -45,12 +49,16 @@ async function navigate(page) {
    if (renderFn) {
      content.innerHTML = '';
      content.className = 'page-content page-enter';
+      if (bar) bar.style.width = '70%';
      await renderFn();
+      if (bar) { bar.style.width = '100%'; setTimeout(() => { bar.style.width = '0'; bar.classList.remove('active'); }, 400); }
    } else {
      content.innerHTML = `<div class="empty-state"><div class="empty-state-icon">🔍</div><div class="empty-state-title">Page not found</div><div class="empty-state-desc">The page "${hash}" doesn't have a render function</div></div>`;
+      if (bar) { bar.style.width = '0'; bar.classList.remove('active'); }
    }
  } catch (err) {
    content.innerHTML = `<div class="empty-state"><div class="empty-state-icon">⚠</div><div class="empty-state-title">Failed to load</div><div class="empty-state-desc">${escapeHtml(err.message)}</div><button class="btn btn-primary mt-3" onclick="navigate('dashboard')">Go to Dashboard</button></div>`;
+    if (bar) { bar.style.width = '0'; bar.classList.remove('active'); }
  }
 }

--- a/dashboard/index.html
+++ b/dashboard/index.html
@ -11,9 +11,10 @@
 <script src="https://cdn.jsdelivr.net/npm/chart.js@4.4.7/dist/chart.umd.min.js"></script>
 </head>
 <body>
+  <div id="topLoadingBar"></div>
  <nav id="sidebar" class="sidebar">
    <div class="sidebar-header">
-      <div class="logo">
+      <div class="logo" onclick="navigate('dashboard')" style="cursor:pointer" title="Go to Dashboard">
        <svg class="logo-svg" viewBox="0 0 36 36" width="36" height="36">
          <defs><linearGradient id="lg" x1="0%" y1="0%" x2="100%" y2="100%"><stop offset="0%" stop-color="#6c5ce7"/><stop offset="100%" stop-color="#fd79a8"/></linearGradient></defs>
          <rect x="2" y="2" width="32" height="32" rx="10" fill="url(#lg)" opacity="0.15"/>
--- a/dashboard/pages/chat.js
+++ b/dashboard/pages/chat.js
@ -129,7 +129,11 @@ async function sendChatMessage() {
  const typingId = showTypingIndicator(agent);

  try {
-    const r = await api.chat(agent, message);
+    // Client-side timeout: 200s (slightly more than Hermes' 180s backend timeout)
+    const controller = new AbortController();
+    const timeoutId = setTimeout(() => controller.abort(), 200000);
+    const r = await api.chat(agent, message, controller);
+    clearTimeout(timeoutId);
    removeTypingIndicator(typingId);
    addChatMessage('assistant', r.response.content, agent);

@ -138,7 +142,8 @@ async function sendChatMessage() {
    window._chatHistory.push({ role: 'assistant', content: r.response.content, agent });
  } catch (err) {
    removeTypingIndicator(typingId);
-    addChatMessage('assistant', `⚠ Error: ${err.message}`, agent);
+    const msg = err.name === 'AbortError' ? 'Request timed out after 200s' : err.message;
+    addChatMessage('assistant', `⚠ Error: ${msg}`, agent);
  }
 }

--- a/dashboard/pages/prompts.js
+++ b/dashboard/pages/prompts.js
@ -47,17 +47,20 @@ async function viewPrompt(name) {

  const displayName = name.replace(/-/g, ' ').replace(/\b\w/g, c => c.toUpperCase());

+  // Store raw content for clipboard copy (avoids HTML entity encoding issue)
+  window._promptCopyContent = content;
+
  showModal(`Prompt: ${displayName}`, `
    <pre style="white-space:pre-wrap;font-size:12px;max-height:60vh;overflow:auto">${escapeHtml(content)}</pre>
  `, `
-    <button class="btn btn-primary" onclick="copyPromptContent('${escapeHtml(content).replace(/'/g, "\\'")}')">📋 Copy</button>
+    <button class="btn btn-primary" onclick="copyPromptFromModal()">📋 Copy</button>
    <button class="btn btn-ghost" onclick="closeModal()">Close</button>
  `);
 }

-function copyPromptContent(text) {
-  const decoded = text.replace(/\\'/g, "'");
-  navigator.clipboard.writeText(decoded).then(() => {
+function copyPromptFromModal() {
+  const text = window._promptCopyContent || '';
+  navigator.clipboard.writeText(text).then(() => {
    showToast('Copied to clipboard', 'success');
  }).catch(() => {
    showToast('Failed to copy', 'error');
--- a/dashboard/pages/skills.js
+++ b/dashboard/pages/skills.js
@ -37,7 +37,8 @@ function renderSkillGrid(skills) {
    const lastScore = s.scores && s.scores.length > 0 ? s.scores[s.scores.length - 1] : null;
    const avg = lastScore && lastScore.criteria_scores ? (lastScore.criteria_scores.reduce((a, b) => a + b, 0) / lastScore.criteria_scores.length) : null;
    const icons = ['⚡', '🔧', '📝', '🔍', '🔄', '🎯', '📊', '🛠', '💡', '🧪', '📋', '💾', '💰', '🔄', '🎨'];
-    const icon = icons[Math.floor(Math.random() * icons.length)];
+    const iconIdx = s.name.split('').reduce((a, c) => a + c.charCodeAt(0), 0) % icons.length;
+    const icon = icons[iconIdx];
    return `<div class="skill-card" onclick="showSkillDetail('${s.name}')">
      <div class="skill-card-header">
        <div class="skill-card-icon">${icon}</div>
--- a/dashboard/styles.css
+++ b/dashboard/styles.css
@ -1,14 +1,14 @@
 :root {
-  --bg-primary: #0a0e1a;
-  --bg-secondary: #111627;
-  --bg-card: #1a1f3a;
-  --bg-card-hover: #222845;
-  --bg-input: #0d1225;
-  --border: #2a3158;
-  --border-hover: #3d46a0;
-  --text-primary: #e8eaf0;
-  --text-secondary: #8b90b0;
-  --text-muted: #5a6080;
+  --bg-primary: #0d1117;
+  --bg-secondary: #151b24;
+  --bg-card: #1c2333;
+  --bg-card-hover: #232b3e;
+  --bg-input: #0d1117;
+  --border: #293042;
+  --border-hover: #4a5280;
+  --text-primary: #e6edf3;
+  --text-secondary: #8b949e;
+  --text-muted: #5d6570;
  --accent: #6c5ce7;
  --accent-light: #8b7cf7;
  --accent-glow: rgba(108, 92, 231, 0.25);
@ -57,7 +57,7 @@ body {
  background: var(--bg-primary);
  color: var(--text-primary);
  display: flex;
-  min-height: 100vh;
+  height: 100vh;
  overflow: hidden;
  -webkit-font-smoothing: antialiased;
 }
@ -81,7 +81,7 @@ body {
  top: 0;
  z-index: 100;
  transition: var(--transition);
-  overflow: hidden;
+  box-shadow: 2px 0 24px rgba(0,0,0,0.08);
 }

 .sidebar.collapsed {
@ -92,7 +92,7 @@ body {
 .sidebar.collapsed .nav-label,
 .sidebar.collapsed .sidebar-section-label,
 .sidebar.collapsed .agent-status-text,
-.sidebar.collapsed .theme-label { display: none; }
+.sidebar.collapsed .nav-badge { display: none; }
 .sidebar.collapsed .nav-item {
  justify-content: center;
  padding: 10px 0;
@ -105,9 +105,9 @@ body {
 .sidebar.collapsed .nav-item.active::before { left: -6px; }
 .sidebar.collapsed .sidebar-section { padding: 4px 0; }
 .sidebar.collapsed .sidebar-header { justify-content: center; padding: 16px 0; gap: 0; }
-.sidebar.collapsed .sidebar-toggle { position: absolute; right: -14px; top: 18px; background: var(--bg-secondary); border: 1px solid var(--border); border-left: none; border-radius: 0 var(--radius-sm) var(--radius-sm) 0; width: 20px; height: 32px; opacity: 0.8; z-index: 10; }
-.sidebar.collapsed .sidebar-toggle:hover { opacity: 1; }
-.sidebar.collapsed .sidebar-footer { justify-content: center; padding: 12px 0; flex-direction: column; gap: 8px; }
+.sidebar.collapsed .sidebar-toggle { position: absolute; right: -18px; top: 18px; background: var(--bg-secondary); border: 1px solid var(--border); border-left: none; border-radius: 0 var(--radius-sm) var(--radius-sm) 0; width: 24px; height: 36px; opacity: 0.9; z-index: 200; cursor: pointer; box-shadow: 3px 0 8px rgba(0,0,0,0.15); transition: var(--transition); display: flex; align-items: center; justify-content: center; }
+.sidebar.collapsed .sidebar-toggle:hover { opacity: 1; width: 28px; }
+.sidebar.collapsed .sidebar-footer { justify-content: center; padding: 12px 0; flex-direction: column; gap: 8px; align-items: center; }

 .sidebar-header {
  display: flex;
@ -166,8 +166,10 @@ body {
 .nav-icon { font-size: 16px; width: 22px; text-align: center; flex-shrink: 0; }
 .nav-badge {
  margin-left: auto; font-size: 10px; font-weight: 700;
-  padding: 2px 6px; border-radius: 10px; 
+  padding: 2px 7px; border-radius: 10px; 
  background: var(--accent-glow); color: var(--accent-light);
+  line-height: 1.2; white-space: nowrap; min-width: 18px;
+  text-align: center; flex-shrink: 0;
 }

 .sidebar-footer {
@ -176,22 +178,35 @@ body {
 }
 .agent-status-bar { display: flex; align-items: center; gap: 8px; font-size: 12px; }
 .agent-dot { width: 8px; height: 8px; border-radius: 50%; flex-shrink: 0; }
-.agent-dot.online { background: var(--green); box-shadow: 0 0 8px var(--green-dim); }
+.agent-dot.online { background: var(--green); box-shadow: 0 0 8px var(--green-dim); animation: pulseDot 2s ease-in-out infinite; }
 .agent-dot.offline { background: var(--red); box-shadow: 0 0 8px var(--red-dim); }
-.agent-dot.warning { background: var(--yellow); box-shadow: 0 0 8px var(--yellow-dim); }
+.agent-dot.warning { background: var(--yellow); box-shadow: 0 0 8px var(--yellow-dim); animation: pulseDot 3s ease-in-out infinite; }
+
+@keyframes pulseDot {
+  0%, 100% { box-shadow: 0 0 4px var(--green-dim); }
+  50% { box-shadow: 0 0 16px var(--green-dim), 0 0 32px var(--green-dim); }
+}
+@keyframes pulseDotWarning {
+  0%, 100% { box-shadow: 0 0 4px var(--yellow-dim); }
+  50% { box-shadow: 0 0 16px var(--yellow-dim), 0 0 32px var(--yellow-dim); }
+}
+.agent-dot.warning { animation: pulseDotWarning 3s ease-in-out infinite; }
 .agent-status-text { color: var(--text-muted); font-size: 11px; white-space: nowrap; }
 .theme-btn {
  background: var(--bg-card); border: 1px solid var(--border);
  border-radius: var(--radius-sm); cursor: pointer;
  width: 30px; height: 30px; display: flex; align-items: center; justify-content: center;
  font-size: 14px; transition: var(--transition);
+  flex-shrink: 0;
 }
 .theme-btn:hover { background: var(--bg-card-hover); border-color: var(--border-hover); }
+.sidebar.collapsed .theme-btn { width: 34px; height: 34px; font-size: 16px; border-color: var(--accent-glow); }

 /* ─── Main Content ─── */
 .main-content {
  flex: 1; display: flex; flex-direction: column;
  height: 100vh; overflow: hidden;
+  min-width: 0;
 }

 .topbar {
@ -219,7 +234,9 @@ body {
 .topbar-search input::placeholder { color: var(--text-muted); }

 .page-content {
-  flex: 1; overflow-y: auto; padding: 24px 28px;
+  flex: 1; display: flex; flex-direction: column;
+  overflow-y: auto; padding: 24px 28px;
+  min-height: 0;
 }

 /* ─── Page Header ─── */
@ -228,7 +245,7 @@ body {
  margin-bottom: 24px;
 }
 .page-header-left {}
-.page-title { font-size: 22px; font-weight: 700; }
+.page-title { font-size: 22px; font-weight: 700; background: linear-gradient(135deg, var(--text-primary), var(--accent-light)); -webkit-background-clip: text; -webkit-text-fill-color: transparent; background-clip: text; }
 .page-subtitle { font-size: 13px; color: var(--text-secondary); margin-top: 2px; }

 /* ─── Cards ─── */
@ -247,9 +264,10 @@ body {
 }
 .card:hover {
  border-color: var(--border-hover);
-  transform: translateY(-1px);
-  box-shadow: 0 8px 32px rgba(0,0,0,0.12);
+  transform: translateY(-2px);
+  box-shadow: 0 12px 40px rgba(0,0,0,0.15);
 }
+.card:active { transform: translateY(0); }
 .card-header {
  display: flex; align-items: center; justify-content: space-between;
  margin-bottom: 16px;
@ -259,7 +277,6 @@ body {
 .stat-card {
  position: relative; overflow: hidden;
  padding: 20px;
-  backdrop-filter: blur(4px);
 }
 .stat-card::after {
  content: ''; position: absolute; top: -20%; right: -10%;
@ -574,6 +591,22 @@ pre {
 }
 .agent-card:hover { border-color: var(--border-hover); }

+/* ─── Top Loading Bar ─── */
+#topLoadingBar {
+  position: fixed; top: 0; left: 0; width: 0; height: 2px;
+  background: linear-gradient(90deg, var(--accent), var(--pink), var(--accent));
+  z-index: 99999;
+  transition: width 0.3s cubic-bezier(0.4, 0, 0.2, 1), opacity 0.3s;
+  opacity: 0;
+  background-size: 200% 100%;
+  animation: loadingBarShimmer 1s linear infinite;
+}
+#topLoadingBar.active { opacity: 1; }
+@keyframes loadingBarShimmer {
+  0% { background-position: 200% 0; }
+  100% { background-position: -200% 0; }
+}
+
 /* ─── Page Transitions ─── */
 .page-enter {
  animation: pageIn 0.3s cubic-bezier(0.4, 0, 0.2, 1);
@ -656,7 +689,8 @@ pre {
 /* ─── Chat Interface ─── */
 .chat-layout {
  display: flex; gap: 0;
-  height: calc(100vh - 140px);
+  flex: 1;
+  min-height: 0;
  background: var(--bg-card);
  border: 1px solid var(--border);
  border-radius: var(--radius-lg);
@ -785,5 +819,4 @@ pre {
@media (max-width: 700px) {
  .chat-sidebar { display: none; }
  .chat-message { max-width: 95%; }
-  .chat-layout { height: calc(100vh - 180px); }
 }
--- a/data/chat-history.json
+++ b/data/chat-history.json
@ -1,60 +0,0 @@
-{
-  "messages": [
-    {
-      "id": "7ab2a689",
-      "role": "user",
-      "agent": "opencode",
-      "content": "hello",
-      "timestamp": "2026-05-17T10:36:36.669215+00:00"
-    },
-    {
-      "id": "08585d4a",
-      "role": "assistant",
-      "agent": "opencode",
-      "content": "Command timed out after 30 seconds",
-      "timestamp": "2026-05-17T10:37:06.778402+00:00"
-    },
-    {
-      "id": "c4f995e1",
-      "role": "user",
-      "agent": "opencode",
-      "content": "hello",
-      "timestamp": "2026-05-17T10:40:43.973330+00:00"
-    },
-    {
-      "id": "70a6b91d",
-      "role": "assistant",
-      "agent": "opencode",
-      "content": "Hey Mihir. Ready when you are.\n",
-      "timestamp": "2026-05-17T10:40:56.421491+00:00"
-    },
-    {
-      "id": "dfe1a0c3",
-      "role": "user",
-      "agent": "opencode",
-      "content": "hello",
-      "timestamp": "2026-05-17T10:42:24.521156+00:00"
-    },
-    {
-      "id": "13f79ad0",
-      "role": "assistant",
-      "agent": "opencode",
-      "content": "\u23f1 Agent 'opencode' took too long to respond.\n\nYour message has been dispatched. For a full response, run the agent CLI directly in your terminal.\n\n**Message:** hello",
-      "timestamp": "2026-05-17T10:42:36.569740+00:00"
-    },
-    {
-      "id": "0801ef96",
-      "role": "user",
-      "agent": "opencode",
-      "content": "hey",
-      "timestamp": "2026-05-17T10:49:09.907635+00:00"
-    },
-    {
-      "id": "13259f25",
-      "role": "assistant",
-      "agent": "opencode",
-      "content": "Hey Mihir. What are we working on today?\n\u001b[0m\n> build \u00b7 deepseek-v4-flash-free\n\u001b[0m\n",
-      "timestamp": "2026-05-17T10:49:19.652896+00:00"
-    }
-  ]
-}
--- a/data/cost-history.json
+++ b/data/cost-history.json
@ -1,6 +0,0 @@
-{
-  "entries": [],
-  "daily_totals": {},
-  "monthly_projection": 0,
-  "free_tier_alerts": []
-}
--- a/server.py
+++ b/server.py
@ -21,7 +21,17 @@ from fastapi.responses import HTMLResponse, JSONResponse, Response
 from fastapi.staticfiles import StaticFiles
 from pydantic import BaseModel

-app = FastAPI(title="Agentic OS", version="1.0.0")
+app = FastAPI(title="Agentic OS", version="1.1.0")
+
+# Load OpenRouter API key from Hermes .env
+HERMES_ENV = Path.home() / ".hermes" / ".env"
+if HERMES_ENV.exists():
+    for line in HERMES_ENV.read_text().splitlines():
+        line = line.strip()
+        if line and not line.startswith("#") and "=" in line:
+            k, v = line.split("=", 1)
+            if k == "OPENROUTER_API_KEY":
+                os.environ[k] = v  # last value wins (matches shell sourcing)

 # CORS for local dev
 app.add_middleware(
@ -85,34 +95,28 @@ def append_audit(entry: dict):
    with open(audit_file, "a") as f:
        f.write(json.dumps(entry) + "\n")

-# ─── Agent Discovery (with cache) ─────────────────────────────────
-
-_agent_cache = {}
-_agent_cache_time = 0
+# ─── Agent Discovery (instant filesystem checks) ────────────────────

 def check_agent(name: str) -> dict:
-    global _agent_cache, _agent_cache_time
-    now = time.time()
-    if now - _agent_cache_time < 15 and name in _agent_cache:
-        return _agent_cache[name]
+    """Instant filesystem-based check. No subprocess needed."""
    try:
        if name == "opencode":
-            r = subprocess.run(["opencode", "--version"], capture_output=True, text=True, timeout=3)
-            status = "online" if r.returncode == 0 else "error"
+            exists = shutil.which("opencode") is not None
+            status = "online" if exists else "offline"
        elif name == "hermes":
-            r = subprocess.run(["hermes", "--version"], capture_output=True, text=True, timeout=3)
-            status = "online" if r.returncode == 0 else "error"
+            exists = shutil.which("hermes") is not None
+            status = "online" if exists else "offline"
        elif name == "gemini":
-            r = subprocess.run(["gemini", "--version"], capture_output=True, text=True, timeout=3)
-            status = "online" if r.returncode == 0 else "error"
+            # Gemini has valid OAuth tokens logged in
+            oauth = Path.home() / ".gemini" / "oauth_creds.json"
+            exists = shutil.which("gemini") is not None
+            logged_in = oauth.exists() and "ya29" in oauth.read_text()
+            status = "online" if exists and logged_in else "offline" if not exists else "warning"
        else:
            status = "offline"
    except Exception:
        status = "offline"
-    result = {"name": name, "status": status}
-    _agent_cache[name] = result
-    _agent_cache_time = now
-    return result
+    return {"name": name, "status": status}

 # ─── Routes: Status ───────────────────────────────────────────────

@ -445,37 +449,105 @@ def save_chat_message(msg: dict):
        history["messages"] = history["messages"][-200:]
    CHAT_HISTORY_FILE.write_text(json.dumps(history, indent=2))

+def run_cli(args: list, timeout: int = 30) -> tuple:
+    r = subprocess.run(args, capture_output=True, text=True, timeout=timeout)
+    return r.returncode, r.stdout, r.stderr
+
+def clean_hermes_output(raw: str) -> str:
+    """Strip CLI metadata from Hermes output, returning only the AI response."""
+    if not raw:
+        return ""
+    lines = raw.split('\n')
+    in_box = False
+    content_lines = []
+    for line in lines:
+        if '╭─' in line:
+            in_box = True
+            continue
+        if '╰─' in line:
+            in_box = False
+            continue
+        if in_box:
+            # Remove ANSI escape codes and leading whitespace
+            cleaned = line.strip()
+            if cleaned:
+                content_lines.append(cleaned)
+    if content_lines:
+        return '\n'.join(content_lines)
+    # Fallback: if no box found, return last non-metadata line
+    non_meta = [l.strip() for l in lines if l.strip() and not l.startswith(('Query:', 'Initializing', '──', 'Resume', 'Session:', 'Duration:', 'Messages:'))]
+    return '\n'.join(non_meta[-5:]) or raw
+
 def execute_agent(agent: str, message: str) -> str:
    try:
        if agent == "opencode":
-            r = subprocess.run(
-                ["opencode", "run", message],
-                capture_output=True, text=True, timeout=12
-            )
-            out = (r.stdout or "") + (r.stderr or "")
-            if r.returncode == 0:
-                return out if out.strip() else f"✓ opencode received your message.\n\n**Message:** {message}\n\n> opencode is processing this in its interactive session. For a full response, run `opencode run \"{message[:60]}\"` in your terminal."
-            return out or f"opencode returned exit code {r.returncode}"
+            try:
+                code, out, err = run_cli(["opencode", "run", "--format", "json", message], timeout=30)
+            except subprocess.TimeoutExpired:
+                return f"⏱ Agent 'opencode' timed out.\n\nOpenCode's model is taking too long. Try running `opencode run \"{message[:60]}\"` directly in your terminal.\n\n**Message:** {message[:100]}"
+            if code == 0:
+                response_text = ""
+                for line in (out or "").split('\n'):
+                    line = line.strip()
+                    if not line:
+                        continue
+                    try:
+                        event = json.loads(line)
+                        if event.get("type") == "text":
+                            text = event.get("part", {}).get("text", "")
+                            if text:
+                                response_text += text + "\n"
+                    except (json.JSONDecodeError, KeyError):
+                        continue
+                if response_text:
+                    return response_text.strip()
+                return f"**opencode**\n\nProcessed your message.\n\n**Message:** {message[:100]}"
+            err_msg = (err or "").strip()
+            return err_msg or f"opencode returned exit code {code}"
+
        elif agent == "hermes":
-            r = subprocess.run(
-                ["hermes", "--message", message],
-                capture_output=True, text=True, timeout=12
-            )
-            out = (r.stdout or "") + (r.stderr or "")
-            return out if out.strip() else f"✓ Dispatched to Hermes.\n\n**Message:** {message}"
+            try:
+                code, out, err = run_cli(["hermes", "chat", "-q", message], timeout=180)
+            except subprocess.TimeoutExpired:
+                return f"⏱ Hermes timed out.\n\nThe model took too long to respond. Try a shorter query or check your OpenRouter rate limits.\n\n**Message:** {message[:100]}"
+            if code == 0:
+                cleaned = clean_hermes_output(out or "")
+                if cleaned:
+                    return cleaned
+                # Empty response from model - return useful fallback
+                return f"**Hermes**\n\nReceived your message but the model returned an empty response. Try rephrasing your query.\n\n**Message:** {message}"
+            err_msg = (err or "").strip()
+            if "invalid choice" in err_msg or "usage:" in err_msg:
+                return f"**Hermes needs setup**\n\nRun `hermes setup` or check your config.\n\n**Details:** {err_msg[:200]}"
+            return err_msg or f"hermes returned exit code {code}"
+
        elif agent == "gemini":
-            r = subprocess.run(
-                ["gemini", "ask", message],
-                capture_output=True, text=True, timeout=30
-            )
-            out = (r.stdout or "") + (r.stderr or "")
-            return out if out.strip() else f"✓ Sent to Gemini CLI.\n\n**Message:** {message}"
+            for attempt, (args, to) in enumerate([
+                (["-y", "-m", "gemini-2.5-flash"], 60),
+                (["-y"], 40),
+            ]):
+                try:
+                    code, out, err = run_cli(["gemini", *args, message], timeout=to)
+                except subprocess.TimeoutExpired:
+                    if attempt == 0:
+                        continue
+                    return f"⏱ Gemini timed out.\n\nTry running `gemini \"{message[:60]}\"` directly.\n\n**Message:** {message[:100]}"
+                if code == 0:
+                    return (out or "").strip() or f"**Gemini CLI**\n\nProcessed your query.\n\n**Message:** {message}"
+                err_msg = (err or "").strip()
+                if attempt == 0 and ("model" in err_msg.lower() or "not found" in err_msg.lower()):
+                    continue
+                if "auth" in err_msg.lower() or "login" in err_msg.lower():
+                    return f"**Gemini needs re-auth**\n\nRun `gemini auth login` to re-authenticate.\n\n**Details:** {err_msg[:200]}"
+                return err_msg or f"gemini returned exit code {code}"
+            return "Gemini CLI did not return a response."
+
        else:
            return f"Unknown agent: {agent}"
    except subprocess.TimeoutExpired:
-        return f"⏱ Agent '{agent}' took too long to respond.\n\nYour message has been dispatched. For a full response, run the agent CLI directly in your terminal.\n\n**Message:** {message[:100]}"
+        return f"⏱ Agent '{agent}' timed out.\n\nRun `{agent} --help` in your terminal for CLI usage.\n\n**Message:** {message[:100]}"
    except FileNotFoundError:
-        return f"⚠ Agent '{agent}' CLI is not installed. Run `pip install {agent}` or install it manually."
+        return f"⚠ Agent '{agent}' CLI not installed. Install it and try again."
    except Exception as e:
        return f"⚠ Error communicating with {agent}: {str(e)}"