claude-view plugin
A dashboard for every Claude Code session on the machine: live agents, cost, history search and MCP tools for Claude.
- 124 ★ on the repository
- Last commit
Extended access
Why this level
- The dashboard reads all Claude Code session logs, including prompts and tool calls.
- Session sharing by link moves its data off the machine, although the link is encrypted.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add plugins/claude-viewDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The dashboard reads all Claude Code session logs, including prompts and tool calls.
- Session sharing by link moves its data off the machine, although the link is encrypted.
This is third-party code. Review the repository files before installing.
What it does
claude-view is a local dashboard that watches every Claude Code session on your computer. Cards show the last message, model, cost and status. A context fill gauge and a prompt-cache countdown flag sessions about to run out of room. Subagents appear as a tree with cost and tool calls, and hooks and MCP calls show inline in the conversation. The `@claude-view/plugin` plugin exposes dashboard data to Claude through 85 MCP tools and nine skills, including `/session-recap`, `/daily-cost` and `/standup`, and starts the dashboard at session start. The backend is written in Rust, the binary weighs about 10 MB, and search runs over raw JSONL logs.
Who it is for. People who keep several Claude Code sessions open at once and want token spend and agent state in one window.
Good fit when
- Ten agents run in parallel and you must see which one finished or awaits approval.
- You need cost per session and project with a token breakdown by model.
- You search for a past decision by message and tool-call text.
- You build a multi-session work summary for a standup with `/standup`.
Not a fit when
- You work on Linux or Windows: the README badge lists macOS.
- You only need a terminal spend report: `ccusage` is enough.
- Your agents are not Claude Code: the dashboard reads this tool's sessions.
Example request
Show how much I spent today by session and which agents are running now.Limitations
The README header lists macOS as the platform. Official builds (`npx claude-view`, the install script) send anonymous usage analytics without session content by default; `CLAUDE_VIEW_TELEMETRY=0` turns it off, and a source build sends nothing. Encrypted session sharing works only when you trigger it. The Insights section and AI Fluency Score are marked experimental. Session phase classification needs a local model with an OpenAI-compatible endpoint.
How to disable. Uninstall the plugin with `claude plugin uninstall claude-view` and stop the dashboard. Delete the binary installed by the script by hand.
FAQ
How do I start the dashboard?
With `npx claude-view` or the install script from the README. The plugin starts the dashboard at every Claude Code session start.
What do the plugin skills do?
`/session-recap` recaps a session, `/daily-cost` shows the day's spend, `/standup` builds a work log, `/search` searches sessions in plain language, and `/export-data` exports CSV and JSON.
Does data leave the machine?
Sessions, code, prompts and paths stay local. The exception is encrypted-link sharing, which the user triggers.
How do I turn analytics off?
Set `CLAUDE_VIEW_TELEMETRY=0` or use the toggle in Settings. A source build sends no analytics.
Can I open a file in my editor?
Yes. The dashboard detects VS Code, Cursor, Zed and other editors and shows an open button for files in the changes view.
Discussion
No questions or reviews yet.
Related
PluginsA local real-time dashboard for Claude Code and Codex sessions: agent activity and tool calls
A local dashboard for Claude Code sessions: timelines, token costs, tools, agents, skills and live sessions
An environment for Claude Code that keeps context across sessions via hooks, ledgers and handoffs, orchestrates agents and saves tokens
A top-style terminal dashboard for Claude Code and Codex sessions: spend, tokens, context load, processes and tool calls.
A Claude Code plugin that draws a 3D office where your session and subagents work at desks from real transcripts, zero config.
Local search across your whole Claude Code session history: decisions, commands, variable values and code you already wrote, through 13 MCP tools
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Repository stars | Main difference |
|---|---|---|---|
| claude-view (this card) | no | 124 | |
| AI Observer | no | 283 | AI Observer collects tokens and latency across several agents; claude-view is built for Claude Code and adds a live feed, search and an MCP plugin |
| Claude Code Agent Monitor | no | 1,145 | Agent Monitor shows Claude Code and Codex sessions in real time; claude-view adds analytics, history and MCP tools |
| ccusage | no | 19,337 | ccusage prints spend reports from logs; claude-view shows live sessions, subagent trees and search |
README in short
The dashboard indexes Claude Code logs with no setup and shows them as live cards, history and analytics. The plugin adds access to the same data from inside Claude itself.