Now liveThe Skillselion MCP - thousands of ranked skills, loaded into your agent mid-task. No install.Get it →
Claudicle avatar

Claudicle

  • 6 repo stars
  • Updated April 13, 2026
  • telepenin/claudicle

Configure Claude Code telemetry for Claudicle

About

claudicle is a Claude Code skill in the AI & Agent Building category. Configure Claude Code telemetry for Claudicle

  • claudicle
  • AI & Agent Building
  • AI-coding skill

Claudicle by the numbers

  • Data as of Aug 1, 2026 (Skillselion catalog sync)
/plugin marketplace add telepenin/claudicle
/plugin install claudicle@claudicle

Add your badge

Show developers this skill is listed on Skillselion. Paste this into your README.

Listed on Skillselion
repo stars6
Last updatedApril 13, 2026
Repositorytelepenin/claudicle

What it does

Configure Claude Code telemetry for Claudicle

README.md

Claudicle

CI codecov npm version License: MIT

Claudicle

The chronicles of Claude. Collect and visualize your Claude Code session telemetry — costs, tokens, tool usage, and full conversation transcripts — in a self-hosted web UI.

Why Claudicle?

Claude Code doesn't show you where your money goes. Claudicle gives you a dashboard for API costs, token trends, and tool performance across sessions. It also captures full conversation transcripts so you can review, search, and export any session — whether you're tracking your own usage or monitoring a team.

Installation Guide | Configuration | Contributing

Quick Start

git clone https://github.com/telepenin/claudicle && cd claudicle
cp .env.example .env
docker compose up -d

Configure Claude Code to send telemetry — use the plugin or add the settings manually:

Option A: Claude Code plugin (recommended)

Install the plugin with /plugin marketplace add telepenin/claudicle, then run /claudicle-configure. It writes the required env vars and interactively sets resource attributes (project, developer, team).

Option B: Manual

Add to ~/.claude/settings.json:

{
  "env": {
    "CLAUDE_CODE_ENABLE_TELEMETRY": "1",
    "OTEL_LOGS_EXPORTER": "otlp",
    "OTEL_METRICS_EXPORTER": "otlp",
    "OTEL_EXPORTER_OTLP_PROTOCOL": "http/protobuf",
    "OTEL_EXPORTER_OTLP_ENDPOINT": "http://localhost:4318",
    "OTEL_RESOURCE_ATTRIBUTES": "project=my-project,developer=your-name"
  }
}

Open http://localhost:3000. For full setup details, see the Installation Guide.

Alternative: Install via npm

If you already have ClickHouse running:

npm install -g claudicle

# Install the UI (saves credentials, initializes schema, registers as system service)
claudicle setup ui --user claude --password claude

# Install the OTel Collector on every machine running Claude Code
claudicle setup collector --user claude --password claude

Features

  • Dashboard — cost and token trends, top models and tools, events by type, filterable by project/environment/team/developer
  • Session browser — searchable list with message counts, subagent and error indicators, date range filtering
  • Session detail view — rendered conversations with rich tool visualizations (diffs, code blocks, search results, MCP tools, nested subagent sessions), live tail for active sessions
  • Session export — portable .tar.gz archives preserving the ~/.claude/projects/ structure, restore and resume on any machine

Architecture

                   ┌──OTLP HTTP──▶  OTel Collector (:4318) ──────────────────────────────┐
Claude Code ───────┤                                                                       ├──▶  ClickHouse  ──▶  Next.js App (:3000)
                   └──JSONL──▶  ~/.claude/projects/*.jsonl  ──▶  OTel Collector (filelog) ┘
  • OTel Collector (otelcol-contrib) — receives OTLP on port 4318 + tails JSONL files, exports to ClickHouse
  • ClickHouse (Docker) — stores events, metrics, and session logs
  • Next.js App (Docker) — web UI and API routes

What Gets Collected

Source Data
OTel events user prompts, tool results (name/duration/success), API requests (model/cost/tokens), errors, tool decisions
OTel metrics token usage, cost, session count, lines of code, commits, PRs, active time
JSONL logs (optional) full conversation transcripts — prompts, responses, thinking blocks, tool I/O

Run the OTel Collector locally to enable JSONL log collection. See the Installation Guide.

Roadmap

  • Session analysis reports — per-session health dashboards with error timelines, tool success rates, and actionable insights
  • Cross-session A/B testing — compare metrics across skill versions, CLAUDE.md configs, and MCP setups
  • Self-improving skills — analyze session history to auto-improve Claude Code skills (superpowers): detect failure patterns (over-tasked subagents, skill-triggered errors, MCP misfires), generate targeted SKILL.md edits, and A/B test skill versions across sessions to measure impact
  • Custom OTel Collector — fork otelcol-contrib with a filelog receiver that only watches JSONL sessions for projects with telemetry enabled (currently it watches all ~/.claude/projects/**/*.jsonl)
  • JSONL redaction — strip sensitive data before ingestion
  • OpenClaw support

Tech Stack

Next.js 16 (App Router) · ClickHouse · OTel Collector · Tailwind CSS · shadcn/ui · recharts

Contributing

Contributions welcome! See CONTRIBUTING.md for development setup and guidelines.

License

MIT

Related skills

This week in AI coding

Five minutes, every Monday - the tools, releases and tactics for developers.

unsubscribe anytime.