Feature library
AI coding analytics features
Explore vibestats feature pages for Claude Code stats, Codex CLI stats, wrapped summaries, activity heatmaps, cost tracking, and privacy-first AI coding analytics.
Feature
Claude Code stats
Track Claude Code usage locally with daily, monthly, session, model, token, and cost reports.
- Daily, monthly, total, session, and model views
- Token and estimated cost reporting
- Works with wrapped pages and activity heatmaps
Feature
Codex CLI stats
Track OpenAI Codex CLI usage locally with token, cost, session, model, and date-based reports.
- Dedicated Codex source mode
- Daily, monthly, total, session, and model reporting
- Fits side-by-side with combined and wrapped views
Feature
Combined AI coding stats
Merge Claude-compatible and Codex CLI usage into one view for daily reporting, wrapped summaries, and activity analysis.
- One combined scope across supported local sources
- Useful for retrospectives and weekly reporting
- Pairs well with activity and wrapped pages
Feature
AI coding wrapped
Create a wrapped-style annual recap for Claude Code and Codex CLI usage with totals, streaks, and model mix.
- Year-style recap with totals and streaks
- Readable for both solo devs and teams
- Backed by local aggregate data instead of prompts
Feature
AI coding activity heatmap
Generate GitHub-style activity heatmaps for Claude Code and Codex CLI usage with streaks and rolling totals.
- GitHub-style calendar view
- Recent totals and streaks in the same surface
- Useful for weekly and monthly retrospectives
Feature
Token and cost tracking
Track input, output, cache, total token volume, and estimated cost across AI coding sessions.
- Input, output, cache read, and cache write visibility
- Estimated cost alongside usage volume
- Works across source-specific and combined views
Feature
Model breakdown
Inspect which Claude-compatible or Codex models dominated your usage and how their token share changed over time.
- Group usage by model instead of date
- Useful for stack reviews and experimentation
- Pairs well with wrapped summaries and cost views
Feature
Session breakdown
Inspect AI coding usage per session so you can separate heavy sessions from routine daily activity.
- Per-session reporting
- Useful for debugging spikes and outliers
- Supports main-session vs subagent-aware workflows
Feature
AI coding share pages
Publish aggregate usage, wrapped, and activity pages without exposing raw prompts or responses.
- Hosted aggregate share pages
- Supports usage, wrapped, and activity reports
- Better communication surface than raw command output
Feature
Privacy-first analytics
Use local-first AI coding analytics without sending prompts, responses, or project paths to a hosted dashboard.
- Local-first reporting model
- Aggregate shares instead of raw transcripts
- Fits privacy-sensitive development workflows
Feature
Claude diagnostics
Inspect local Claude account state, cache freshness, and usage-data limit signals without external API access.
- Reads local Claude state only when requested
- Useful for debugging and limit visibility
- Complements session and usage reports
Feature
JSON and automation exports
Generate machine-readable vibestats output for dashboards, automations, scripts, and downstream analytics.
- Machine-readable JSON output
- Fits shell scripts and downstream systems
- Useful for custom dashboards and weekly reports