GitHub - Dicoangelo/claude-command-center: Real-time analytics dashboard for Claude Code: sessions, routing, costs, model performance. SSE streaming, auto-refresh.

A self-healing, real-time analytics dashboard that monitors every heartbeat of your Claude Code infrastructure.

Dashboard Preview

_{All 17 tabs cycling automatically — Overview through Velocity}

📸 Tab Gallery — Click to browse all 17 tabs

1. Overview

2. Memory

3. Activity

4. Cost

5. Projects

6. Commands

7. Routing

8. Co-Evolution

9. Context Packs

10. Session Outcomes

11. Productivity

12. Tool Analytics

13. Supermemory

14. Cognitive

15. Infrastructure

16. Autonomy

17. Velocity

Live Metrics

┌────────────────────────────────────────────────────────────────────────────────┐
│                                                                                │
│   4,211        382K          89.9K       $8.8K        99.8%       17          │
│   SESSIONS     MESSAGES      TOOL CALLS  API VALUE    CACHE HIT   TABS        │
│                                                                                │
│   ══════════════════════════════════════════════════════════════════════════   │
│                                                                                │
│   • SSE streams every 3 seconds       • Self-heals on failure                 │
│   • SQLite single source of truth     • 8 LaunchAgent daemons                 │
│   • Zero-downtime reloads             • Autonomous watchdog                   │
│                                                                                │
└────────────────────────────────────────────────────────────────────────────────┘

System Architecture

%%{init: {'theme': 'dark', 'themeVariables': { 'primaryColor': '#00ff88', 'primaryTextColor': '#fff', 'primaryBorderColor': '#00ff88', 'lineColor': '#00ff88', 'secondaryColor': '#0a1628', 'tertiaryColor': '#0d1117', 'clusterBkg': '#0d1117', 'clusterBorder': '#00ff88'}}}%%
flowchart TB
    subgraph CCC["⚡ CLAUDE COMMAND CENTER ⚡"]
        direction TB

        subgraph PRESENTATION["🖥️ PRESENTATION LAYER"]
            direction LR
            BROWSER["Browser Dashboard\n━━━━━━━━━━\n15 Interactive Tabs\nChart.js Visualizations"]
            SSE["SSE Live Stream\n━━━━━━━━━━\n3s Real-Time Updates\nAuto-Reconnect"]
        end

        subgraph SERVER["🔌 SERVER LAYER"]
            direction LR
            API["API Server\n━━━━━━━━━━\nThreadingHTTPServer\n12 REST Endpoints"]
            STREAMER["DataStreamer\n━━━━━━━━━━\nMD5 Change Detection\nBroadcast to Clients"]
        end

        subgraph INTELLIGENCE["🧠 INTELLIGENCE LAYER"]
            direction LR
            BRAIN["Autonomous Brain\n━━━━━━━━━━\nPattern Detection\nAnomaly Alerts"]
            INTEL["Intelligence Layer\n━━━━━━━━━━\nLeverage Points\nOptimization"]
            HEAL["Self-Heal Engine\n━━━━━━━━━━\n8 Error Patterns\n89% Recovery Rate"]
        end

        subgraph DATA["💾 DATA LAYER"]
            direction LR
            SQLITE[("SQLite\nclaude.db\n━━━━━━━━━━\nSessions • Stats\nRouting • Events")]
            PIPELINE["Data Pipeline\n━━━━━━━━━━\nTranscript Scanner\nToken Aggregator"]
        end
    end

    OPERATOR(("👤 OPERATOR"))

    OPERATOR <==>|"http://localhost:8766"| BROWSER
    BROWSER <-->|"EventSource"| SSE
    SSE <--> API
    API <--> STREAMER
    STREAMER --> SQLITE
    API --> SQLITE
    BRAIN --> SQLITE
    INTEL --> BRAIN
    HEAL -.->|"auto-fix"| PIPELINE
    PIPELINE -->|"sync"| SQLITE
    HEAL -.->|"monitor"| API

    subgraph DAEMONS["🔧 LAUNCHAGENT DAEMONS"]
        WATCHDOG["Watchdog\n━━━━━━━━━━\nAlways-On Guardian"]
        BOOTSTRAP["Bootstrap\n━━━━━━━━━━\nAuto-Start"]
    end

    WATCHDOG -.->|"keepalive"| API
    WATCHDOG -.->|"monitor"| HEAL
    BOOTSTRAP -.->|"init"| DAEMONS

    style CCC fill:#0d1117,stroke:#00ff88,stroke-width:3px
    style PRESENTATION fill:#0a1628,stroke:#00ff88,stroke-width:2px
    style SERVER fill:#0a1628,stroke:#00d9ff,stroke-width:2px
    style INTELLIGENCE fill:#0a1628,stroke:#ff6b6b,stroke-width:2px
    style DATA fill:#0a1628,stroke:#ffd700,stroke-width:2px
    style DAEMONS fill:#0a1628,stroke:#9945ff,stroke-width:2px
    style OPERATOR fill:#00ff88,stroke:#fff,stroke-width:2px,color:#0d1117

_{🔄 Self-healing architecture — Every layer monitors the one below it}

Dashboard Tabs

📊 Overview

Mission Control

Sessions, messages, tool calls, ROI multiplier, 10x Power Dashboard, activity charts, token breakdown, hourly heatmap.

Real-Time SSE Chart.js

💰 Cost Analytics

ROI Tracking

API value vs $200/mo subscription, per-model token-level cost breakdown, cache efficiency (99.8%), daily cost trends, ROI calculator.

$8.8K Value 21.5x ROI $8.6K Saved

🧠 Cognitive OS

Energy-Aware Intelligence

Current mode, energy level, flow state, session fate prediction, weekly energy map, flow state history.

Flow State Fate Prediction Energy Map

🛡️ Infrastructure

Self-Healing Systems

System health, 8 daemon statuses, watchdog activity, self-heal runs, recovery rate, protection layers.

8 Daemons 89% Recovery Auto-Heal

🔀 Routing

DQ Score Intelligence

Model routing accuracy, DQ score distribution, cost savings from smart routing, expertise-aware decisions.

DQ: 0.665 52.5% Savings 1.2K Queries

🔧 Tool Analytics

Usage Intelligence

89.9K total tool calls, 20 unique tools, success rates, daily averages, top tool breakdown.

89.9K Calls 100% Success 20 Tools

Full Tab Registry

#	Tab	Key Metrics	Visual
1	Overview	4.2K sessions, 382K messages, 89.9K tools	Charts + Cards
2	Memory	88 knowledge items, 98 unique tags	Donut + Bars
3	Activity	Daily messages, session heatmap	Time Series
4	Cost	$8.8K API value, 21.5x ROI, $200/mo sub	Stacked + Gauge
5	Projects	19 repos, 10.6K commits, 3.9M LOC	Table + Stats
6	Commands	65 commands, 10 categories	Frequency
7	Routing	1.2K queries, DQ 0.665, 52.5% savings	Distribution
8	Co-Evolution	289 mods, self-improvement tracking	Timeline
9	Context Packs	8 packs, 3.7K tokens	Pack Cards
10	Session Outcomes	2.8K analyzed, 55.6% success	Pie + Trends
11	Productivity	38.5% score, 17.3K writes, 27.6K reads	Bar + Line
12	Tool Analytics	89.9K calls, 20 tools, 100% success	Bar + Table
13	Supermemory	5K items, 871 learnings, 152 reviews due	Spaced Rep
14	Cognitive	Flow state, fate prediction, energy map	Line + Heatmap
15	Infrastructure	8/8 daemons, 1.2K heals, 89% recovery	Status Grid
16	Autonomy	Streak detection, longest runs, tool patterns	Timeline + Bars
17	Velocity	10D vector field, VAAC signals, composition	Radar + Line

Data Pipeline

%%{init: {'theme': 'dark', 'themeVariables': { 'primaryColor': '#00ff88', 'primaryTextColor': '#fff', 'lineColor': '#00d9ff', 'secondaryColor': '#0a1628', 'tertiaryColor': '#0d1117'}}}%%
flowchart LR
    T["📄 Session\nTranscripts"] --> FIX["🔧 fix-all-dashboard-data.py\nScan + Aggregate"]
    FIX --> SYNC["🔄 sqlite-to-jsonl-sync.py\nBridge Layer"]
    SYNC --> SQL["🗄️ ccc-sql-data.py\nExport JSON"]
    SQL --> GEN["⚙️ ccc-generator.sh\nRender 505K HTML"]
    GEN --> API["🚀 ccc-api-server.py\nServe Live + SSE"]
    API --> BROWSER["🖥️ Browser\nReal-Time Dashboard"]

    style T fill:#1a1a2e,stroke:#00ff88,color:#fff
    style FIX fill:#1a1a2e,stroke:#00d9ff,color:#fff
    style SYNC fill:#1a1a2e,stroke:#00d9ff,color:#fff
    style SQL fill:#1a1a2e,stroke:#00d9ff,color:#fff
    style GEN fill:#1a1a2e,stroke:#ffd700,color:#fff
    style API fill:#1a1a2e,stroke:#00ff88,color:#fff
    style BROWSER fill:#00ff88,stroke:#fff,color:#0d1117

_{📡 Transcript to dashboard in 5 pipeline stages — SQLite is the single source of truth}

Self-Healing Infrastructure

The system that fixes itself before you notice anything is broken

┌──────────────────────────────────────────────────────────────────────────────────┐
│                         SELF-HEALING ARCHITECTURE                                  │
├──────────────────────────────────────────────────────────────────────────────────┤
│                                                                                   │
│  ┌─────────────┐    ┌─────────────┐    ┌─────────────┐    ┌─────────────┐       │
│  │  WATCHDOG   │───▶│  SELF-HEAL  │───▶│  BOOTSTRAP  │───▶│  AUTOPILOT  │       │
│  │             │    │   ENGINE    │    │             │    │             │       │
│  │ Always-On   │    │ 8 Patterns  │    │ Daemon Mgmt │    │ Autonomous  │       │
│  │ 60s Checks  │    │ Auto-Fix    │    │ LaunchAgent │    │ Operations  │       │
│  └─────────────┘    └─────────────┘    └─────────────┘    └─────────────┘       │
│                                                                                   │
│  ┌────────────────────────────────────────────────────────────────────────────┐  │
│  │                         PROTECTION LAYERS                                  │  │
│  │                                                                             │  │
│  │   L1: Watchdog ──monitor──▶ L2: KeepAlive ──restart──▶ L3: Bootstrap      │  │
│  │   L4: Wake Hook ──trigger──▶ L5: Self-Heal ──deep scan every 6h──▶ Fix    │  │
│  │                                                                             │  │
│  │        1,195 Runs    •    89% Recovery    •    3 Auto-Healed (24h)        │  │
│  └────────────────────────────────────────────────────────────────────────────┘  │
│                                                                                   │
└──────────────────────────────────────────────────────────────────────────────────┘

Layer	Component	Interval	Capability
1	Watchdog	60s	Monitors all daemons, restarts failures
2	KeepAlive	launchd	OS-level auto-restart via LaunchAgent
3	Bootstrap	Login	Ensures entire infrastructure is alive
4	Wake Hook	Sleep	Triggers bootstrap after system wake
5	Self-Heal	6h	Deep health scan, 8 error patterns, auto-fix

SSE Streaming Protocol

⚡ REAL-TIME LAYER

Feature	Detail
`POLLING`	3-second SQLite polling via DataStreamer
`DETECTION`	MD5 hash change detection
`BROADCAST`	Multi-client SSE broadcast
`RECONNECT`	Exponential backoff, 50 max retries

🖥️ DASHBOARD LAYER

Feature	Detail
`INDICATOR`	Green LIVE / Yellow CONNECTING / Red OFFLINE
`DOM PATCH`	Real-time stat card updates
`TIMESTAMP`	Last update time in LIVE badge
`FALLBACK`	Static mode when served from file://

Setup

Full installation from a fresh clone

Prerequisites: Python 3.12+, macOS (for LaunchAgent), SQLite 3

# 1. Clone
git clone https://github.com/Dicoangelo/claude-command-center.git
cd claude-command-center

# 2. Install dev dependencies
python3 -m venv .venv && source .venv/bin/activate
pip install -e '.[dev]'

# 3. Create symlinks (scripts → ~/.claude/scripts/)
for f in scripts/ccc-*.py scripts/ccc-*.sh; do
  ln -sf "$(pwd)/$f" ~/.claude/scripts/$(basename "$f")
done

# 4. Initialize the database
sqlite3 ~/.claude/data/claude.db < config/schema.sql

# 5. Generate the dashboard
make generate

# 6. Start the server
make serve
# → http://localhost:8766/dashboard

# 7. (Optional) Install LaunchAgent for always-on mode
cp deploy/com.claude.api-server.plist ~/Library/LaunchAgents/
launchctl load ~/Library/LaunchAgents/com.claude.api-server.plist

Quick Start

# Live dashboard (syncs data + starts server + opens browser)
ccc-live

# Static dashboard (generates HTML file)
ccc

# Server only
python3 scripts/ccc-api-server.py --port 8766

# Health check
curl http://localhost:8766/api/health

Development

make help        # Show all targets
make test        # Run test suite (138 tests)
make lint        # Lint with ruff
make typecheck   # Type check with mypy
make backup      # Backup SQLite database
make all         # Lint + format + typecheck + test

API Endpoints

GET  /dashboard           # Live dashboard with injected SSE client
GET  /api/stream          # SSE event stream (real-time updates)
GET  /api/stats           # Session stats and usage data
GET  /api/cost            # Cost data and savings
GET  /api/routing         # Routing metrics and DQ scores
GET  /api/sessions        # Session outcomes
GET  /api/tools           # Tool usage statistics
GET  /api/git             # Git activity metrics
GET  /api/health          # System health check
GET  /api/fate            # Cognitive fate predictions
GET  /api/cognitive       # Cognitive OS state
GET  /api/memory/stats    # Memory store statistics
POST /api/memory/query    # Query across all memory stores

Infrastructure Commands

# Self-heal engine (8 error patterns, auto-fix)
python3 scripts/ccc-self-heal.py

# Watchdog guardian (60s health checks)
python3 scripts/ccc-watchdog.py

# Bootstrap all daemons
bash scripts/ccc-bootstrap.sh

# Quick status check
bash scripts/ccc-status.sh

# Autonomous brain (pattern detection)
python3 scripts/ccc-autonomous-brain.py

# Intelligence layer (leverage points)
python3 scripts/ccc-intelligence-layer.py

File Registry

Layer	File	Size	Purpose
🖥️	`dashboard/claude-command-center.html`	505K	17-tab generated dashboard
🚀	`scripts/ccc-api-server.py`	26K	Live server + SSE streaming
⚙️	`scripts/ccc-generator.sh`	78K	Static HTML generator
🗄️	`scripts/ccc-sql-data.py`	23K	SQLite → JSON exporter
🔧	`scripts/fix-all-dashboard-data.py`	34K	Transcript scanner + aggregator
🔄	`scripts/sqlite-to-jsonl-sync.py`	5K	SQLite ↔ JSONL bridge
📊	`scripts/integrate-untracked-data.py`	12K	Untracked data integration
📈	`scripts/dashboard-sql-loader.py`	12K	SQL data loader
🧠	`scripts/ccc-autonomous-brain.py`	26K	Pattern detection + anomalies
💡	`scripts/ccc-intelligence-layer.py`	20K	Leverage point analysis
🛡️	`scripts/ccc-self-heal.py`	36K	Self-healing engine (8 patterns)
👁️	`scripts/ccc-watchdog.py`	5K	Always-on guardian
🤖	`scripts/ccc-autopilot.py`	6K	Autonomous operations
🔀	`scripts/ccc-migrate-to-sqlite.py`	9K	Migration tooling
🚀	`scripts/ccc-bootstrap.sh`	1K	Daemon bootstrap
📋	`scripts/ccc-status.sh`	3K	Quick health check
💾	`config/datastore.py`	16K	SQLite ORM layer
📐	`config/schema.sql`	14K	Database schema (20 tables, 42 indexes, 6 views)
💿	`scripts/ccc-backup.py`	3K	SQLite backup with rotation
🔧	`deploy/com.claude.api-server.plist`	<1K	LaunchAgent daemon config
🧪	`tests/`	12K	138 tests (API, SQL data, self-heal, datastore, backup)

Deploy

# 1. Copy LaunchAgent plist
cp deploy/com.claude.api-server.plist ~/Library/LaunchAgents/

# 2. Load the daemon (auto-starts, auto-restarts)
launchctl load ~/Library/LaunchAgents/com.claude.api-server.plist

# 3. Verify
curl http://localhost:8766/api/health
# → {"status": "healthy", "daemons_count": 12}

Vision

╔══════════════════════════════════════════════════════════════════════════════╗
║                                                                              ║
║   "I am no longer needed for the actual technical work.                     ║
║    I describe what I want built, in plain English,                          ║
║    and it just... appears."                                                 ║
║                                                                              ║
║   ════════════════════════════════════════════════════════════════════════   ║
║                                                                              ║
║   THE COMMAND CENTER EXISTS BECAUSE:                                          ║
║                                                                              ║
║   ▸ 4,211 sessions generate data that deserves visibility                  ║
║   ▸ Self-healing means never waking up to a broken system                  ║
║   ▸ Real-time streaming replaces manual dashboard refreshes                ║
║   ▸ Every token, every tool call, every session — tracked                  ║
║                                                                              ║
║   ════════════════════════════════════════════════════════════════════════   ║
║                                                                              ║
║   "The system monitors itself so you don't have to."                        ║
║                                                                              ║
╚══════════════════════════════════════════════════════════════════════════════╝

Troubleshooting

Server won't start

Port 8766 may be in use. Kill the stale process and retry:

lsof -ti:8766 | xargs kill -9
python3 scripts/ccc-api-server.py

Dashboard shows stale data

Regenerate from SQLite:

make fix-data    # Scan transcripts → SQLite
make generate    # Regenerate HTML from SQLite

SSE not connecting (LIVE indicator stays yellow/red)

Verify the server is running: curl http://localhost:8766/api/health
Check browser console for CORS errors
SSE only works over HTTP — opening the HTML file directly (file://) shows STATIC mode

ROI numbers look wrong

Verify pricing matches Anthropic's current rates:

python3 -c "from pricing import PRICING; print(PRICING)"
# Should show: opus input=$5/MTok, output=$25/MTok, cache=$0.50/MTok

If wrong, update ~/.claude/config/pricing.py.

LaunchAgent not starting

# Unload and reload
launchctl unload ~/Library/LaunchAgents/com.claude.api-server.plist
launchctl load ~/Library/LaunchAgents/com.claude.api-server.plist

# Check logs
tail -20 /tmp/ccc-api-server.log

Database locked

SQLite WAL mode supports concurrent reads but only one writer. Check for stale connections:

fuser ~/.claude/data/claude.db 2>/dev/null
# Kill stale processes if needed

🤖 Build Log (click to expand)

Date	Action	Outcome
2026-02-14	Engineering hardening	42 tests, CI/CD, pyproject.toml, schema.sql, Makefile, type hints, backup script
2026-02-14	ROI fix + UI polish	Cost tab pricing fix, live /api/cost, nav overflow, text clamp
2026-02-13	Cost model overhaul	Token-level pricing ($200/mo sub), symlink consolidation, README accuracy
2026-02-12	Live SSE streaming	Real-time dashboard with 3s updates, LIVE indicator
2026-02-12	Token pipeline fix	Backfilled daily_stats from sessions table (79 rows)
2026-02-12	Recovery data fix	Mapped SQLite Row objects to proper JS format
2026-02-12	17-tab audit	All tabs verified clean (zero undefined/NaN/Invalid Date)
2026-02-12	Private repo created	20 files, 16,613 lines committed
2026-01-31	Self-healing engine	8 error patterns, 89% recovery rate
2026-01-29	Autonomous brain	Pattern detection + anomaly alerts
2026-01-21	SQLite migration	Replaced JSONL files with unified database
2026-01-19	Intelligence layer	Leverage point analysis
2026-01-17	17-tab dashboard	Full analytics suite with Chart.js
2026-01-14	Initial API server	REST endpoints on port 8766
2026-01-09	Watchdog + Bootstrap	LaunchAgent daemon infrastructure

Architecture: Built entirely with Claude Code (Opus 4.6) — sovereign AI infrastructure monitoring sovereign AI infrastructure.

Part of the Antigravity Ecosystem | Built by @dicoangelo

Name		Name	Last commit message	Last commit date
Latest commit History 35 Commits
.beads		.beads
.github		.github
.ralph-tui		.ralph-tui
assets		assets
config		config
dashboard		dashboard
deploy		deploy
scripts		scripts
tasks		tasks
tests		tests
.editorconfig		.editorconfig
.gitattributes		.gitattributes
.gitignore		.gitignore
AGENTS.md		AGENTS.md
CCC-PRD-AUDIT.md		CCC-PRD-AUDIT.md
CHANGELOG.md		CHANGELOG.md
LICENSE		LICENSE
Makefile		Makefile
README.md		README.md
pyproject.toml		pyproject.toml

Uh oh!

Folders and files

Latest commit

History

Repository files navigation

Dashboard Preview

📸 Tab Gallery — Click to browse all 17 tabs

1. Overview

2. Memory

3. Activity

4. Cost

5. Projects

6. Commands

7. Routing

8. Co-Evolution

9. Context Packs

10. Session Outcomes

11. Productivity

12. Tool Analytics

13. Supermemory

14. Cognitive

15. Infrastructure

16. Autonomy

17. Velocity

Live Metrics

System Architecture

Dashboard Tabs

📊 Overview

💰 Cost Analytics

🧠 Cognitive OS

🛡️ Infrastructure

🔀 Routing

🔧 Tool Analytics

Full Tab Registry

Data Pipeline

Self-Healing Infrastructure

SSE Streaming Protocol

⚡ REAL-TIME LAYER

🖥️ DASHBOARD LAYER

Setup

Quick Start

Development

API Endpoints

Infrastructure Commands

File Registry

Deploy

Vision

Troubleshooting

About

Topics

Resources

License

Uh oh!

Stars

Watchers

Forks

Releases

Sponsor this project

Uh oh!

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages