Skip to content
Iria Monitor
Iria Monitor

Make AI programming governable in your company.

Iria Monitor gives you the evidence you need to govern it.

What code reaches git, how much survives in production, and how to compare tools, teams and vendors against the same criteria.

If you buy code

Reward the vendors that deliver lasting value.

When you can see which vendors ship code that survives in production, you can renew with confidence, scale what works, and reinvest the savings into more ambitious projects.

Compare vendors on the same KPIs. Grow the partnerships that produce real value.

Vendor scorecard — Q1 2026

Stellar's code lasts. Renew, expand the contract, and use what you learn to set the bar for everyone else.

Stellar Engineering91% survives

74% AI code · 757 lines · 7 commits

Nexus Digital72% survives

82% AI code · 455 lines · 5 commits

Ironforge Systems88% survives

22% AI code · 75 lines · 4 commits

If you deliver code

Show the value you create — and charge for it.

Your team uses AI to ship more. Now you can prove the quality stays. Walk into the renewal conversation with adoption, durability, and churn already on the table — the numbers that justify your rate.

Turn shipped value into a quarterly report your client trusts. Win renewals. Win expansion.

Quality report — ready to export

The report nobody else can produce.

  • AI-assisted code68%
  • Durability at 30 days93%
  • Churn rate4%
  • AI code in production2,340
  • Primary agentclaude-code

If you write code with AI

Get better at the craft. Make every prompt count.

See which agents and prompts produce code that survives. Double down on what works, refine what doesn't, and watch your durability climb month over month.

Your data. Private. On your machine. The honest mirror that helps you grow.

Your personal stats — private

Only you see this

Not your prompts for tests — you can improve that.

  • My AI code this week62%
  • My durability88% ↑ 79%
  • Claude for auth code95%
  • Cursor for tests60%

How it works

01
Install the CLI

Hooks into Claude Code, Cursor, Codex, Windsurf. Installs git hooks. Zero config.

02
Code and push

Use any AI agent. Commit and push normally. Attribution data travels with your code via git notes — an open standard.

03
See the data

In your terminal or on the web. Blame, stats, durability. By vendor, by team, by developer.

iria-monitor stats

AI: 68% · Durability: 91%

Works with your AI coding tools

Claude CodeCursorCodexWindsurf

Compatible with git-ai v3.0.0 standard. Open format. Your data is portable.

INCLUDED IN THE PACKAGE

The complete AI engineering toolkit.

Per-line attribution is the core. But Iria Monitor ships with everything else you need to run AI coding as a managed program — usage, cost, live activity, weekly trends.

LLM usage & cost proxy

Capture every prompt, token, and dollar across Claude, Cursor, Codex, and ChatGPT — attributed to the developer, repo, and branch where it happened. Cost meets value, in one place.

System tray

Live AI % indicator in your system tray on macOS, Windows, and Linux. Click to see today's tokens, cost, durability, and current session — without opening a browser.

Live SSE dashboard

Watch the work happen in real time: which agent is editing which file, which session is burning the most tokens, what's being committed right now.

Weekly burndown & reports

Trend lines, weekly summaries, exportable HTML reports. The artifact you drop into a stand-up, a vendor review, or a board pack.

All included. Free CLI for individuals. Install from this environment's toolkit.

Personal AI Usage Coach

AI is an amplifier — and a mirror.

A local-first mirror that surfaces what your AI is amplifying — no telemetry, no ranking, no shame. Personal DORA in seven signals.

See the coach

Start measuring.

Free CLI, no account needed. Web dashboard for teams.