Why we built a checkpoint for coding agents.
Coding agents are good at logic and indifferent to side effects. The analytics call sitting inside a checkout function is a side effect. So it gets dropped, and your funnel goes quiet for three sprints. Skene exists because someone has to be aware that the call matters.
The pattern we kept seeing.
We watched teams ship faster with Cursor and Claude Code, and we watched their dashboards drift weeks later. The two things were connected. A function got rewritten, the posthog.capture call inside it did not survive, the funnel kept drawing from worse data.
The agents were not careless. They are pattern-matchers operating on logic, and instrumentation is structurally invisible to logic. Skene reads the same code the agent does, but specifically for the calls the agent does not see.
Three co-founders in Helsinki.
€800K pre-seed in 2025. Building Skene full-time.
Teemu Kinos
Co-founder
Built and operated growth at SaaS companies for a decade. Skene is what he wishes existed when the dashboards stopped agreeing with each other.
[LinkedIn]Michele Boggia
Co-founder
PhD in physics. NLP since the start of the field. Handles how Skene reads code at scale.
[LinkedIn]Teppo Hudsson
Co-founder
Built multiple products end to end. Super technical, obsessively product-focused.
[LinkedIn]Covered by 17 outlets across 3 languages.
Independent coverage of Skene's €800K pre-seed round.
Separation of duties
The agent writes the tracking. It should not grade its own work.
Skene does not write code, and it does not compete with Cursor, Claude Code, Codex, or Devin. It runs alongside them. An agent checking the instrumentation it just wrote is the conflict of interest the check exists to remove. The gate has to sit outside the tool that made the change.
It is also not a product analytics tool, an instrumentation library, or a customer data platform. It validates that the events your app writes to Supabase still fire and still match your schema after the next PR. That position, the independent gate in the pull request and the record of what every event is supposed to do, is the point. It is the seat, not the data.
Broken instrumentation is more than a bug.
When dashboards drift, decisions drift. Product reads the funnel wrong. Growth optimizes against bad numbers. An exec makes a roadmap call against a chart that has been broken since a refactor three sprints ago.
The bug is not the missed event. The bug is that the team stopped being able to trust their own data.
