Skip to content
Now in public beta

See every agent.
Every token. Every decision.

The observability layer for Claude Code. Monitor agents, track spend, approve actions — from your browser or phone.

Join the early access waitlist

Get Started Free GitHub
$ curl -fsSL https://get.claudeview.ai/install.sh | sh copy
localhost:47892
Dashboard
Sessions
Search
Patterns
Settings
Active Agents
3
Tokens Today
142,847
Sessions
23
Est. Cost
$4.21
claude-view/main
feature-auth Running — editing src/auth.ts 48,291 tok
docs-update Idle 71,452 tok
api-server/feat/auth
test-runner Waiting for approval 23,104 tok

Your agents, your rules.

Write a plan. Approve steps. Agents build. You sleep. Don't let your $200/mo AI plan sit idle — leverage agents as your 24/7 workforce.

Join the waitlist
implementation-plan.md
Set up project structure
Implement auth module
Write integration tests
Deploy to staging
Approve Reject

Ship features from your phone.

The couch. The train. The beach. Push notification → review → approve → shipped. Your agents keep building while you're away.

Get early access

See everything. Miss nothing.

Every agent. Every token. Every tool call. Real-time dashboards show costs with hover breakdowns, cache efficiency, hook event timelines, sub-agent trees, and worktree branch drift detection — all updated live.

feature-auth $1.42
48,291 tokens · 12 tool calls · 94% cache
test-runner $0.87
23,104 tokens · 8 tool calls · 89% cache
docs-update $2.31
71,452 tokens · 24 tool calls · 91% cache

Full-Text Search

Search across all sessions — messages, tool calls, file paths. Powered by Tantivy.

AI Fluency Score

Track how effectively you collaborate with AI agents over time.

Session Sharing

Share any session via encrypted link. E2E encryption — zero server trust.

100% Local

All data stays on your machine. Zero telemetry. Zero cloud dependency.

Kanban Swimlanes

Group running agents by project or branch. Visual swimlane layout for multi-project workflows.

Cost & Token Breakdown

Hover any session for a detailed token breakdown — input, output, cache read, cache write — with color-coded categories.

142k
Sessions monitored
<1ms
Startup time
15MB
On disk
Zero
Telemetry

Built different from the ground up.

Tool Category Stack Size Live Monitor Search Analytics Session Sharing Kanban View Branch Drift
claude-view Full platform Rust + React 15 MB
opcode Viewer TypeScript ~5 MB
ccusage Cost tracker TypeScript ~3 MB
CodePilot Dashboard Python ~20 MB
claude-run Runner TypeScript ~2 MB

Get started in 10 seconds

No account needed. Just run one command.

Shell (recommended) curl -fsSL https://get.claudeview.ai/install.sh | sh
Plugin claude plugin add @claude-view/plugin
npx npx claude-view

No dependencies for shell install. npx requires Node.js 18+. Works on macOS. Linux coming in v2.1.

Free. Forever. For local use.

Cloud features coming soon.

Individual

Free

$0

See what your Claude Code is actually doing.

  • Session browser & search
  • Cost tracking
  • Community support
Get Started

Pro

$20/mo

Full visibility and control from anywhere.

  • Everything in Free, plus:
  • Unlimited session monitoring
  • Cloud relay — work from any device
  • Mobile app (monitor + control)
  • Remote agent control
  • Plan runner & workflows
  • AI Fluency Score & full analytics
Join Waitlist

Max

$100/mo

For power users running agent fleets.

  • Everything in Pro, plus:
  • 5x usage on all features
  • Agent orchestration engine
  • Quality gates (auto-review)
  • Priority access to new features
Join Waitlist

Team & Enterprise

Team

$30/user/mo

Your whole team, one dashboard.

  • Everything in Max, plus:
  • Team dashboard & shared sessions
  • Usage analytics & reporting
  • Centralized billing
  • Role-based access control
Join Waitlist

Enterprise

Custom

Org-wide deployment with full compliance.

  • Everything in Team, plus:
  • SAML/OIDC SSO
  • SCIM seat management
  • Audit logs & compliance
  • Dedicated support
Contact Sales
See full comparison →

Common questions.

Is claude-view free?
Yes — the Free plan is free forever. Just run curl -fsSL https://get.claudeview.ai/install.sh | sh or npx claude-view. No signup, no API key, no config. Pro and Max plans (coming soon) add cloud relay, mobile app, and agent orchestration.
Does it send data to the cloud?
No. claude-view runs 100% locally by default. All session data stays on your machine. The optional cloud relay (Pro plan) uses end-to-end encryption — your data is never readable by our servers.
Do I need to configure anything?
No. Run curl -fsSL https://get.claudeview.ai/install.sh | sh and it works immediately — zero env files, zero accounts, zero database setup. Or use npx claude-view if you prefer npm. Cloud features like sharing auto-enable when configured, but are completely optional.
What's the difference between plugin and npx?
Three ways to install: Shell (curl -fsSL https://get.claudeview.ai/install.sh | sh) downloads a native binary — fastest, no Node.js needed. Claude Code plugin (claude plugin add @claude-view/plugin) auto-starts claude-view when Claude Code launches, plus adds MCP tools and skills. npx (npx claude-view) runs standalone via npm. All use the same dashboard.
Does it work with Cursor or Windsurf?
claude-view monitors Claude Code sessions specifically — the CLI tool from Anthropic. It reads Claude Code's JSONL conversation logs. Other AI coding tools use different formats and aren't supported.
How much does it cost?
Free forever for local monitoring. Pro ($20/mo) unlocks cloud relay, mobile app, and plan runner. Max ($100/mo) adds 5x usage, agent orchestration, and quality gates. Team ($30/user/mo) and Enterprise (custom) add team features.