Skene is product analytics in your own Supabase that work with any dashboard you use. It adds the tracking you are missing and runs on every pull request, reading your Supabase schema read-only to catch when a code change removes, renames, or breaks the writes that record events, so the data stays complete and the metrics built on it stay trustworthy.

Skene reads the repository and tracks every event your app writes into Supabase. When a pull request changes one of those files, Skene flags removed writes, renamed columns, and payloads your schema cannot store on the PR before the change merges.

Is Skene a coding agent?

No. Skene is not a coding agent and does not compete with Cursor, Claude Code, Codex, or Devin. It runs alongside them as an independent check that validates the event-writing code already in your repository.

What does Skene work with?

Skene reads the events your app writes to Supabase and checks them against your schema. It works alongside coding agents like Cursor, Claude Code, Codex, and Devin, and runs in CI as a GitHub Action.

Skene runs as a GitHub Action that posts a comment on every pull request, with a cloud validation API for programmatic runs and a one-time repository audit to baseline what you already have.

Trust your product data. Live in one click.

Product analytics in your own Supabase, working with any dashboard you use. Skene runs as a GitHub Action on every pull request: it adds the events you're missing and catches a broken one in a PR comment, before the merge and while the data's still recoverable.

Start free

The problem

You can't get today's data back tomorrow.

Early on, analytics is handled, for now. You ship features, your agent rewrites the code that writes the events, and one checkout insert quietly gets renamed in a green PR. Nothing errors. The Supabase table just stops filling, and no one notices for weeks.

Then it bites. You're three slides into a board deck defending an activation number you can't personally verify, and the funnel has gone flat. An engineer burns a sprint and finds the break. You can patch the code tomorrow. You can't get back the weeks of data it was broken. That cohort is gone for good, and the decision already went out the door.

Skene catches that break on the pull request, in the same place you already review code, while it's still a comment and the data is still recoverable. The gate sits outside the agent that wrote the code, reads your schema read-only, and your data stays in your own Supabase. Setup takes under ten minutes.

Analytics live in your own Supabase.

Connect once. Skene adds the tracking you're missing, right into the database you already own.

Connect your repo and Supabase, read-only
Adds the events you're missing, indexed against your schema, and checks each one fits a table
Your data stays in your Supabase, never moved or copied
Checks each PR against its preview branch, via Supabase branching

Start free

Starting journey analysis…

This may take a few minutes. Stay on this page to see live pipeline updates.

Analyzing Schema

Events from codebase

Generating Journey

Discovery

From sign up to account creation

SIGNAL	TYPE	FOUND AT	STATUS
Account Created	Supabase	public.accounts	Can be tracked (DB)
Signup Page Viewed	GitHub	app/admin/(auth)/login/page.tsx	Recommended
Signup Form Submitted	GitHub	app/admin/(auth)/login/page.tsx	Recommended
Auth User Created	Supabase	auth.users	Can be tracked (DB)
auth.users.insert	GitHub	–	Can be tracked (DB)
user creates an account record	GitHub	–	Can be tracked (DB)

refactor: consolidate checkout flow tracking

Open

feat/checkout-v2→main

skenebotcommented now

Analytics diff detected

Missing callcheckout_started: not found in new state

Renamed eventpurchase_complete → order_confirmed

Payload shape changecart_viewed: field "items" removed

Conditional no longer firesupsell_shown: condition block unreachable

Caught on review. Not on a dashboard that's been wrong for weeks.

Tracking that doesn't quietly break.

Every way a code change can quietly drop an event. Caught on the pull request that caused it.

Tracking your code stopped recording, or renamed so it no longer lands
Data shaped wrong, so your database silently drops it
Tracking that used to fire but stopped, for some users or some flows
Flagged as a PR comment, not weeks later

Start free

Works with any dashboard. Tells you what pays.

Skene feeds the dashboard you already use, and answers the questions that move revenue.

Point any dashboard or BI tool at your own Supabase
Or ask the questions that move revenue from inside your coding agent, over MCP. No new dashboard to build
Skene gives the answer, not another chart to babysit
Turnkey dollar-revenue viewRoadmap

Start free

Analytics events

checkout_started

purchase_complete

cart_viewed

public·events

event_nametext

order_idint4

cart_itemsmissing

created_attimestamptz

Skene checking events against schema…

feat: add purchase_complete tracking

Open

feat/purchase-events→main

skenebotcommented now

Schema mismatch detected

purchase_complete→public.events

Type mismatchpayload.purchase_id · text → column is int4

Missing columnpayload.cart_items · column does not exist

Numbers you can defend. Checked on every pull request.

One GitHub Action adds the tracking you're missing, then guards it on every pull request. It reads your Supabase schema read-only, so when an agent's refactor breaks a checkout insert, you hear about it in the PR comment, not three weeks later when the funnel goes flat. Then ask your coding agent for the paths and cohorts that pay.

Start free

1. SKENE ADDS THE EVENTS YOU'RE MISSING, INDEXED AGAINST YOUR SUPABASE SCHEMA. ONE ACTION, LIVE IN MINUTES.

2. AN AGENT RENAMES A TRACKING CALL? SKENE CATCHES IT ON YOUR PR. MISS IT, AND THAT DATA IS GONE FOR GOOD.

3. EVERY EVENT IN YOUR FUNNEL STILL FIRES AND LANDS IN THE RIGHT TABLE. THE NUMBER HOLDS.

Four ways to plug Skene in.

Ask questions from your coding agent (MCP server). Run it on every pull request (GitHub Action). Call it from your own scripts (Cloud API). Or start with a one-time check of your repo.

Same engine behind each.

Start free Read more

MCP server

Cursor · Claude Code

GitHub Action

Runs on every PR

Cloud API

Any script, any time

Repo audit

One-time · no commitment

MCP server

Skene runs before the agent commits. Catches analytics issues in the agent loop, not after the PR lands.

skene mcp --cursor

Separation of duties

The agent writes the tracking. It should not grade its own work.

An agent checking the tracking it just wrote is the conflict of interest the check exists to remove. Skene runs alongside Cursor, Claude Code, and Codex, not inside them. The gate sits outside the tool that made the change: it reads the diff against your Supabase schema, sees a write stop landing in its table, and leaves the verdict as a PR comment. The reviewer answers to your schema, not to the agent it is reviewing.

It is not a dashboard you babysit or a copy of your data. Your data stays in your database. Skene owns the gate, and the trusted record you query over MCP for what drives revenue.

119 stars

Open GitHub Start free →

Patch the code tomorrow. The data from this week is already gone.

A renamed event passes every test, ships green, and sits broken for weeks before anyone notices. Skene runs as a GitHub Action on every pull request, reads your Supabase schema read-only, and comments the moment a write breaks, while your data is still recoverable. Usage-based on tokens, with $5 of free cloud credit to start. A PR that touches none of your tracking costs nothing.

Start free

Billed monthlyBilled yearly

ProPopular

$29/mo

1.5M monthly tokens

Local MCP server
GitHub Action with PR comments
Cloud validation API
1.5M monthly tokens

Start free →

Scale

$99/mo

6M monthly tokens

Local MCP server
GitHub Action with PR comments
Cloud validation API
6M monthly tokens (4x Pro)

Start free →

Ultra

$199/mo

15M monthly tokens

Local MCP server
GitHub Action with PR comments
Cloud validation API
15M monthly tokens (10x Pro)
Priority support

Start free →

SIGNAL

TYPE

FOUND AT

STATUS

Account Created

Supabase

public.accounts

Can be tracked (DB)

Signup Page Viewed

GitHub

app/admin/(auth)/login/page.tsx

Recommended

Signup Form Submitted

GitHub

app/admin/(auth)/login/page.tsx

Recommended

Auth User Created

Supabase

auth.users

Can be tracked (DB)

auth.users.insert

GitHub

–

Can be tracked (DB)

user creates an account record

GitHub

–

Can be tracked (DB)

Numbers you can defend. Checked on every pull request.

Patch the code tomorrow. The data from this week is already gone.

Trust your product data. Live in one click.

You can't get today's data back tomorrow.

Analytics live in your own Supabase.

Tracking that doesn't quietly break.

Works with any dashboard. Tells you what pays.

Numbers you can defend. Checked on every pull request.

Four ways to plug Skene in.

The agent writes the tracking. It should not grade its own work.

Built in the open,
with our design partners.

Patch the code tomorrow. The data from this week is already gone.

Catch the break on the PR, or lose the cohort for good.

Trust your product data. Live in one click.

You can't get today's data back tomorrow.

Analytics live in your own Supabase.

Tracking that doesn't quietly break.

Works with any dashboard. Tells you what pays.

Numbers you can defend. Checked on every pull request.

Four ways to plug Skene in.

The agent writes the tracking. It should not grade its own work.

Built in the open,
with our design partners.

Patch the code tomorrow. The data from this week is already gone.

Catch the break on the PR, or lose the cohort for good.

Trust your product data. Live in one click.

You can't get today's data back tomorrow.

Analytics live in your own Supabase.

Tracking that doesn't quietly break.

Works with any dashboard. Tells you what pays.

Numbers you can defend. Checked on every pull request.

Four ways to plug Skene in.

The agent writes the tracking. It should not grade its own work.

Built in the open,with our design partners.

Patch the code tomorrow. The data from this week is already gone.

Catch the break on the PR, or lose the cohort for good.

Trust your product data. Live in one click.

You can't get today's data back tomorrow.

Analytics live in your own Supabase.

Tracking that doesn't quietly break.

Works with any dashboard. Tells you what pays.

Numbers you can defend. Checked on every pull request.

Four ways to plug Skene in.

The agent writes the tracking. It should not grade its own work.

Built in the open,with our design partners.

Patch the code tomorrow. The data from this week is already gone.

Catch the break on the PR, or lose the cohort for good.

Built in the open,
with our design partners.

Built in the open,
with our design partners.