Every session has a score.

Know what your Claude Code sessions actually produce. Qualito scores every session, detects regressions, and tells you what to change — so you get better at working with AI, not just spending on it.

uvx qualito setup
See Your Score → View on GitHub
Terminal
$ qualito setup
  Scanning ~/.claude/projects/ ...
  Found 8 projects (312 sessions, ~$186 total):
    #  Project                Sessions  Est. Cost
    1  my-app                     89    $52.30
    2  api-service                67    $41.20
    3  frontend                   54    $38.10
    ...
  Importing all projects...
  ████████████████████████████ 312/312
  Results:
  ┌─────────────┬──────┬────────┬─────────┐
  │ Workspace   Runs Avg DQI Cost   
  ├─────────────┼──────┼────────┼─────────┤
  │ my-app         89   0.78  $52.30 
  │ api-service     67   0.65  $41.20 
  │ frontend        54   0.82  $38.10 
  └─────────────┴──────┴────────┴─────────┘
  Average DQI: 0.74 (Good)
  Setup complete.
Works with Claude Code
PyPI v0.1.1
MIT License
82 tests passing

Cost tracking tells you what you spent.
Not whether it was worth it.

One developer tracked 147 decision points across 23 sessions. Found 25% of time lost to context re-establishment. Another lost $234 from a single missed magic number.

Cost tracking tells you what you spent. Qualito tells you why — and what to change.

Everything you need to measure delegation quality.

DQI Scoring

Every session, scored 0–100

Your AI sessions are no longer a black box. See which ones delivered and which went in circles.

Experiments

Stop debating. Test it.

Does your 2,500-line CLAUDE.md actually help? Run both approaches for a week. Qualito tells you which one wins — with statistical significance, not vibes.

Incident Detection

Catch regressions before they cost you

Qualito auto-detects when your session quality drops — cost spikes, error loops, declining scores. You get alerted, not surprised.

Pattern Recognition

See what you actually do

Qualito classifies your sessions by task type and surfaces patterns. Find out that your code reviews score 0.84 but your refactors score 0.52.

MCP Integration

Ask Claude about your quality

Add one line to your config. Ask “how am I doing?” Claude analyzes your sessions and tells you — right where you work.

Cross-Project Discovery

All your projects, one command

Qualito scans every Claude Code project on your machine. No per-project setup. Import everything in under 2 minutes.

Test what works. Drop what doesn't.

Run controlled experiments on your AI workflows. Qualito tracks the results with real statistics.

Strategy A: Detailed context WINNER
0.82
Sessions 24
Avg cost $1.40
Success 91%
vs
Strategy B: Minimal prompts
0.61
Sessions 24
Avg cost $3.20
Success 67%

Statistical significance: p = 0.003 (Wilcoxon test)

Switching to Strategy A saves $43/month

Define experiments in one command
Qualito tracks sessions automatically
Get verdicts backed by real statistics, not gut feel

See your quality at a glance.

qualito.ai/dashboard
74 Avg DQI
Runs (30d) 312
Total Cost $186.40
Incidents 0
SLO Status Q A C
DQI Trend (30d)

From scorecard to coach

01

See

Score every session. See where quality drops. Know your cost per run.

Available now
02

Improve

Run experiments. Test your prompting strategies. Apply what works.

Available now
03

Automate

Qualito learns what works for you and suggests changes automatically. Productive friction, not a black box.

Coming Q3 2026

Three ways to start measuring.

1

Cloud Dashboard

Register, copy one command, done. Your sessions sync to qualito.ai where you can track trends across all projects.

Create Account →
2

CLI Wizard

Full control over what to import. Run the setup wizard, pick your projects, and get scores in your terminal.

uvx qualito setup
3

MCP-First

Ask Claude about your quality. Add the MCP server to your editor and get DQI insights inline as you code.

uvx qualito-mcp

Start free. Upgrade when you need more.

Free
$0 forever
  • 100 runs per month
  • 1 workspace
  • DQI scoring
  • CLI + MCP tools
  • 7-day history
  • Community support
Install Free
Pro
$29 / month
  • Unlimited runs
  • Unlimited workspaces
  • Everything in Free
  • Full cloud dashboard
  • Unlimited history
  • Team features
  • Priority support
Start Pro

Common questions.

Your next AI coding session has a score.

See it.

uvx qualito setup
See Your Score →

Free for 100 runs/month. No credit card required.