13days

Claude Panel — Development skill for Claude Code

Development community

⌘ Zero-dependency local dashboard for Claude Code — commands, skills, agents, plugins, workflows, sessions & a ccusage-accurate usage dashboard.

How to install Claude Panel

This entry records only its repository, not the path inside it, so there is no exact command to give. Open 13days/claude-panel and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Claude Panel does

⌘ Zero-dependency local dashboard for Claude Code — commands, skills, agents, plugins, workflows, sessions & a ccusage-accurate usage dashboard.

Alternatives in Development

Ccusage — CLI for analyzing Claude Code/Codex usage from local JSONL files 11.8k ★
Sanitize — Detect and redact PII from text files — 15 categories (SSNs, credit cards, API keys, etc.), zero dependencies 3.1k ★
Claude Swap — Switch between multiple Claude Code accounts, with automatic rate-limit rotation, usage dashboard, and paralle 2k ★

README

⌘ Claude Panel

**See exactly what Claude Code sends the model — and everything else in your `~/.claude`.**

A beautiful, zero-dependency local dashboard for Claude Code: browse and edit your slash commands / skills / agents / plugins / workflows, search and replay past sessions, track token spend — and, uniquely, **capture the real system prompt, tool definitions, and every LLM round-trip** behind each conversation.

npx claude-code-panel

[中文文档](README.zh-CN.md) · [Quick Start](#quick-start) · [Features](#features) · [Screenshots](#screenshots)

![zero dependencies](https://img.shields.io/badge/dependencies-0-brightgreen) ![node >= 18](https://img.shields.io/badge/node-%3E%3D18-blue) ![license MIT](https://img.shields.io/badge/license-MIT-orange) ![i18n](https://img.shields.io/badge/i18n-中文%20%7C%20EN%20%7C%20日本語-purple)

![demo](https://raw.githubusercontent.com/13days/claude-panel/main/docs/demo.gif?v=2)

Why

Claude Code scatters everything across `~/.claude` — commands here, skills there, plugins in a JSON manifest, usage buried in hundreds of MB of transcript JSONL. And the most interesting thing of all — *what actually gets sent to the model* — isn't saved anywhere. Claude Panel fixes both:

Find any command / skill / plugin / past session without digging through the CLI — one searchable place, editable in the browser.
Recover the session you lost after a crash or reboot: full-text search your history, one-click claude --resume.
Understand how a reply was produced: the exact system prompt, tool definitions, full context, and the input/output of every LLM round a query triggered.

The killer feature: see every LLM round-trip

Claude Code transcripts don't store the system prompt or the raw request/response. Claude Panel runs a tiny **local reverse proxy** that Claude Code talks through, records each API call, and folds it back into the session replay — matched to the exact turn.

Open a session, expand a