Skip to main content

Cross-tool usage analytics for Claude Code and Codex CLI

Project description

ai-cli-insights

Cross-tool usage analytics for Claude Code and Codex CLI. Collects session data from both tools and generates rich HTML insight reports with narratives, trend charts, friction analysis, and actionable recommendations.

Features

  • Unified data collection from Claude Code (~/.claude/) and Codex CLI (~/.codex/)
  • Three report modes: cross-tool comparison, Claude-only, Codex-only
  • Period-over-period comparison, trend sparklines, quality scoring
  • Friction pattern analysis with evidence-based recommendations
  • Prompt library with copy-to-clipboard templates
  • Zero third-party dependencies — pure Python standard library

Installation

pip install ai-cli-insights

Quick Start

# Generate a cross-tool report for the last 30 days
ai-cli-insights generate --days 30 --tool all

# Claude-only report
ai-cli-insights generate --tool claude

# Codex-only report
ai-cli-insights generate --tool codex

# Short form (generate is the default command)
ai-cli-insights --days 7 --tool claude

The HTML report is saved to the output directory and can be opened in any browser.

Configuration

# Create a config template
ai-cli-insights init

This writes a config.json to your config directory:

  • macOS: ~/Library/Application Support/ai-cli-insights/config.json
  • Linux: ~/.config/ai-cli-insights/config.json

Edit it to customize:

  • domain_patterns — keyword patterns to classify sessions into work domains
  • project_patterns — patterns to identify which project a session belongs to
  • template_modes — report titles and slugs

Without a config file, reports still work — domains and projects will just show raw paths.

Claude Code Skill

# Install a Claude Code skill for easy /multi-insights access
ai-cli-insights install-skill

This writes a SKILL.md to ~/.claude/skills/multi-insights/.

CLI Reference

ai-cli-insights [command] [options]

Commands:
  generate        Generate an HTML insights report (default)
  init            Write a config template to the config directory
  install-skill   Install Claude Code skill file

Generate options:
  --days N        Analyze the last N days (default: 30)
  --tool MODE     all | claude | codex (default: all)
  --analyst WHO   codex | claude — label the analyzing model (default: codex)
  --output-dir    Override the output directory

python -m ai_cli_insights also works.

Data Sources

Tool Source What it provides
Claude Code ~/.claude/usage-data/session-meta/*.json Session metadata, tool counts, tokens
Claude Code ~/.claude/usage-data/facets/*.json Outcomes, friction, goals
Codex CLI ~/.codex/state_5.sqlite Thread metadata, tokens
Codex CLI ~/.codex/sessions/*.jsonl Event details, timestamps, tools
Codex CLI ~/.codex/history.jsonl User message counts

License

MIT

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

ai_cli_insights-0.1.0.tar.gz (52.0 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

ai_cli_insights-0.1.0-py3-none-any.whl (64.9 kB view details)

Uploaded Python 3

File details

Details for the file ai_cli_insights-0.1.0.tar.gz.

File metadata

  • Download URL: ai_cli_insights-0.1.0.tar.gz
  • Upload date:
  • Size: 52.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.2

File hashes

Hashes for ai_cli_insights-0.1.0.tar.gz
Algorithm Hash digest
SHA256 af90fbba5a8072a7b1440b889e88c4d6d60f542347b3c5a510ffcf0cb454b2c8
MD5 4e44a436d98f6b45511e4daa1ee85d32
BLAKE2b-256 2a392778dd0354e3a0046bdad14d8c1111c2a3ea255d040edda28b1941d62647

See more details on using hashes here.

File details

Details for the file ai_cli_insights-0.1.0-py3-none-any.whl.

File metadata

File hashes

Hashes for ai_cli_insights-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 ecf863ca22c7d0defd28c3b2f2de5f0d9e4de0fe725a2624a3ba40559fad44a0
MD5 403c09f1689eff64466a660ab65bbef6
BLAKE2b-256 11e8414e3920229db094bca5552ac4f8c4f58597bb63ccaf3a1c9e9fa80b4f9b

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page