You shipped
it with AI.
Is it actually
working?

Sypnia watches your app in production. When something breaks, it tells you in plain words what happened and to whom, and hands the fix to your coding agent.

For people who build with Lovable, Bolt, v0, Replit, Cursor or Claude Code.

A customer clicks Pay Checkout page POST /api/checkout POST /v1/payments 400 A customer clicks Pay POST /api/checkout POST /v1/payments 400
New problem2 min agoExample

3 people tried to pay today and could not. The Pay button does nothing since yesterday’s change.

Yesterday’s change renamed the cart’s total. The checkout still asks for the old name, so the payment goes out without an amount and is rejected.

3 people affectedWatch what they sawTrace from click to failed call
Open pull requestCopy prompt for your agent
  • Open sourceEngine in Rust, built in the open.
  • Free to self-hostOne template. Your data stays yours.
  • Cloud from $10 a monthWith a free tier to start.
  • Web, Node, any backendOur SDKs, plus OpenTelemetry for the rest.

It speaks human.

A stack trace tells you where the code stopped. Sypnia tells you who it hurt, what they were trying to do, and what to do about it.

What your app says
ApiError: POST /api/checkout failed
  with 400 (amount_required)
    at createPayment (api.ts:27:11)
    at async confirmPayment (checkout.tsx:142:18)
    at async onClick (pay-button.tsx:18:5)
    at HTMLButtonElement.dispatch
       (react-dom.production.js:3941:16)

request   POST /api/checkout  400
amount    undefined
release   4f2a91c, deployed yesterday 18:40
url       /checkout
events    9      users  3

breadcrumbs
  10:31:28.412  navigation /checkout
  10:31:43.104  ui.click   button#pay
  10:31:45.310  ui.click   button#pay
  10:31:45.502  ui.click   button#pay
  10:31:45.811  ui.click   button#pay
  10:31:46.003  ui.click   button#pay
  10:31:46.974  fetch      POST /api/checkout 400
  10:31:46.981  exception  ApiError
  10:31:58.020  navigation /cart

sessions
  10:31  u_1042     POST /api/checkout 400
  09:47  u_0977     POST /api/checkout 400
  09:11  anonymous  POST /api/checkout 400
and 212 more lines
What Sypnia says Example

3 people tried to pay today and could not. The Pay button does nothing since yesterday’s change.

Who it hit
3 people. Two of them tried again and gave up.
What they were doing
Paying for an order, at the last step of checkout.
Since when
Yesterday’s change, release 4f2a91c, deployed at 18:40. It worked before.
What to do
The checkout still reads the cart’s old field name. Point it at the new one. The fix is ready as a pull request.

It hands the fix to the agent that wrote the code.

You did not write that checkout by hand, and you should not have to debug it by hand. Every problem comes ready for your coding agent, three ways.

  • A prompt, ready to paste

    What happened, where, why, and how to check the fix, written for Cursor, Claude Code or whichever agent you use.

    Prompt for your coding agentExample
    Fix a production bug reported by Sypnia.
    
    What happens: clicking Pay does nothing. 3 people hit it today,
      since release 4f2a91c (yesterday 18:40).
    Where: src/checkout/checkout.tsx:142, confirmPayment()
    Why: that release renamed cart.total to cart.totalCents, but
      checkout.tsx still reads cart.total. The payment goes out
      with amount undefined and the API answers 400.
    Do: read cart.totalCents in confirmPayment(), and show the
      customer a message when a payment is rejected.
    Check: session 8f3a shows the clicks. Add a test that fails
      when the amount is missing.
  • An MCP server your agent can ask

    Connect Sypnia once. From then on your agent can ask what is failing in production and read the trace and the session, as a timeline of events, itself.

    Your coding agent, connected to SypniaExample
    You
    what is failing in production?
    Agent
    sypnia.problems({ status: "open" })
    1 open problem since yesterday’s deploy: 3 people tried to pay and could not. The checkout sends the payment without an amount. I have the trace and the session’s timeline. Want me to fix it?
    You
    yes, and add a test
    Agent
    sypnia.problem({ id: "pay-button-does-nothing" })
    Editing src/checkout/checkout.tsx
  • A pull request with the fix

    On Builder and Team, connect GitHub, turn pull requests on for the project, and Sypnia opens the pull request for you. You read it, you merge it. Nothing ships without you.

    Read the cart total from its new field
    sypnia/pay-button-does-nothingintomainopened by Sypnia, 2 minutes agoExample
    async function confirmPayment(cart) {
    - const payment = await createPayment({ amount: cart.total });
    + const payment = await createPayment({ amount: cart.totalCents });
    + if (!payment.ok) return showPayError(payment);
    return payment;
    }
    Tests passFixes the problem 3 people hit todayReview pull request

Everything it watches, so you can stop watching.

Seven things tell you whether your software and the people using it are well. Sypnia covers them all from one install.

Session replay

See the session as the person lived it: every click, scroll and error, with the moment it broke already marked.

One of the 3 sessionsExample
  1. 0:12Opened checkout
  2. 0:27Clicked Pay
  3. 0:29Clicked Pay 4 more timesRage click
  4. 0:31Payment rejected: no amount
  5. 0:42Left the page
0:31 / 0:421x

Alerts and a daily summary

An alert by e-mail and in the app when it cannot wait, and one short e-mail a day for everything else.

Daily summaryExample

Good morning. Here is yesterday.

  • 1 new problem: 3 people tried to pay and could not, since yesterday’s change. A pull request is waiting for you.
  • The search page got slower: 3.1 seconds to load, it was 1.9.
  • Your app answered all day. 412 people used it.
  • Errors, grouped

    The same failure is one problem, not four hundred emails. Each comes with who it hit, where, and the deploy that caused it.

  • Performance

    Which pages load slowly for real people, and which API calls are slow or failing.

  • Uptime

    Sypnia checks from outside that your app answers, and tells you within about two minutes when it stops.

  • Rage and dead clicks

    Where people click again and again, and the things they click that do nothing.

  • End-to-end traces

    One request followed across your frontend, your API and the services behind it.

It works with what you already built.

Our own light SDKs for the web, Node and Next.js. For any other backend, an open door: if your services speak OpenTelemetry, point them at Sypnia.

  • WebAny framework. Apps in a web view too
  • Node.jsOur SDK
  • Next.jsOur SDK
  • OpenTelemetryPython, Go, PHP, Java, Rust… over OTLP
  • iOS and AndroidNative SDKs planned
  • Python, PHP, Go, RustOwn SDKs planned. OTLP today

One snippet to install

import { init } from '@sypnia/browser';

init({ project: 'YOUR_PROJECT_KEY' });
// errors, replay, performance and clicks are on by default
import { init } from '@sypnia/node';

init({ project: process.env.SYPNIA_PROJECT_KEY });
// requests, outgoing calls, database queries and errors are traced
OTEL_EXPORTER_OTLP_ENDPOINT=https://ingest.sypnia.com/v1/otlp
OTEL_EXPORTER_OTLP_HEADERS=Authorization=Bearer%20YOUR_PROJECT_KEY
OTEL_EXPORTER_OTLP_PROTOCOL=http/protobuf

# opentelemetry-instrument python app.py
OTEL_EXPORTER_OTLP_ENDPOINT=https://ingest.sypnia.com/v1/otlp
OTEL_EXPORTER_OTLP_HEADERS=Authorization=Bearer%20YOUR_PROJECT_KEY
OTEL_EXPORTER_OTLP_PROTOCOL=http/protobuf

# any OpenTelemetry SDK: traces go to Sypnia

One project key everywhere: our SDKs, or any OpenTelemetry exporter.

Open source. Run it yourself, or let us.

The engine that receives, stores and analyzes your data is open. Its core is written in Rust, so it is cheap to run and has room to grow.

Self-host, for free

Deploy the whole engine on your own infrastructure with one template. Your data never leaves your account.

  • The full engine: ingest, storage and analysis
  • One template to deploy
  • Free, with no limits set by us

See the code on GitHub

Sypnia Cloud

If you would rather not run servers, we run it for you. Start on the free tier; paid plans begin at $10 a month.

  • Nothing to deploy or maintain
  • A free tier to start
  • The same engine you can read

Start free

Free to start. From $10 a month when you grow.

Self-hosting is free. On Sypnia Cloud you pay for how much you send, not for features: every plan has replay, traces, the analysis, the prompt and the MCP server.

Plans compared
Self-hosted Free On your own infrastructure Free Free To try it on a real project Builder $10 a month For an app with real users Team $49 a month For several apps and people
Projects No limits set by usProjects, events, replays, people and how long data is kept: as much as your own servers hold.1310
Events a monthAn event is an error, a trace or a performance sample. Clicks and page views are free; in all, up to four times the plan’s events are received. 5,00050,000500,000
Replays a month 505005,000
Longest recording 30 minutes2 hours2 hours
Data kept for 7 days30 days90 days
Replays kept for 7 days30 days30 days
People on the account 1310
Uptime checks 1525
AI analyses a month With your own model key50, at most 10 a day2001,000
Pull requests Sypnia opens a month Not in the engineNot included530
Alert e-mails a month Not in the engine: a webhook1001,0005,000
The prompt and the MCP serverIncludedIncludedIncludedIncluded
SupportCommunityCommunityEmailPriority email

Self-hostedFree

On your own infrastructure

Projects
No limits set by us
Events a month
No limits set by us
Replays a month
No limits set by us
Longest recording
No limits set by us
Data kept for
No limits set by us
Replays kept for
No limits set by us
People on the account
No limits set by us
Uptime checks
No limits set by us
AI analyses a month
With your own model key
Pull requests Sypnia opens a month
Not in the engine
Alert e-mails a month
Not in the engine: a webhook
The prompt and the MCP server
Included
Support
Community

FreeFree

To try it on a real project

Projects
1
Events a month
5,000
Replays a month
50
Longest recording
30 minutes
Data kept for
7 days
Replays kept for
7 days
People on the account
1
Uptime checks
1
AI analyses a month
50, at most 10 a day
Pull requests Sypnia opens a month
Not included
Alert e-mails a month
100
The prompt and the MCP server
Included
Support
Community

Builder$10 a month

For an app with real users

Projects
3
Events a month
50,000
Replays a month
500
Longest recording
2 hours
Data kept for
30 days
Replays kept for
30 days
People on the account
3
Uptime checks
5
AI analyses a month
200
Pull requests Sypnia opens a month
5
Alert e-mails a month
1,000
The prompt and the MCP server
Included
Support
Email

Team$49 a month

For several apps and people

Projects
10
Events a month
500,000
Replays a month
5,000
Longest recording
2 hours
Data kept for
90 days
Replays kept for
30 days
People on the account
10
Uptime checks
25
AI analyses a month
1,000
Pull requests Sypnia opens a month
30
Alert e-mails a month
5,000
The prompt and the MCP server
Included
Support
Priority email

Team grows with you, on the same monthly bill: extra seat $3 extra project $2 +100,000 events $5 +500 replays $5 +100 analyses $3 +10 pull requests $15 +10 uptime checks $1 +2,000 alert e-mails $1 +90 days of data $2. Each a month. Free and Builder grow by moving up a plan.

Prices in US dollars. Self-hosting is always free.

Start free

Honest answers.

Is your question missing? Write to us and a person will answer. support@sypnia.com

How do I start?

Create a project, paste one snippet into your app (or point your OpenTelemetry exporter at Sypnia) and open your app. Its first session shows up, and problems start arriving as sentences.

What does it watch?

The essentials on this page: errors grouped with context, session replay, performance, uptime, rage and dead clicks, end-to-end traces, alerts and a daily summary, plus the loop with your coding agent (the prompt, the MCP server and, on paid plans, GitHub pull requests). If it is not on this page, we are not promising it.

How is it different from Sentry, Datadog or PostHog?

They are excellent and broad, built for engineering teams who are at home in dashboards. Sypnia is narrower on purpose: it is for people who ship with AI and want to be told in a sentence what is wrong, for whom, and how to fix it. It is open source and accepts OpenTelemetry, so you can run it next to another tool or move later.

Where does my data live?

If you self-host, on your own infrastructure and nowhere else. On Sypnia Cloud, on Amazon Web Services in the United States (us-east-1), used only to show you your own app and kept for as long as your plan says.

What does session replay record?

It rebuilds what the page looked like from its changes; it is not a camera on the person. What people type is always masked, and you can mask or leave out any area you mark as sensitive.

Which AI tools does it work with?

The prompt works with any coding agent that accepts text. The MCP server works with agents that support MCP, such as Cursor and Claude Code. Pull requests go through GitHub, on Builder and Team, once you turn them on for a project.

What does it cost?

Self-hosting is free. Sypnia Cloud has a free tier and paid plans from $10 a month.

Is it really open source?

The engine is: ingest, storage and analysis, with its Rust core. You can read it, run it yourself, and keep your data on your own servers.

Sleep. We keep one eye open.

Three steps, and Sypnia is watching your app.

  1. Create a project

    Open a free account and name your app.

  2. Paste the snippet

    One snippet in your app, or point your OpenTelemetry exporter at Sypnia.

  3. See your first session

    Open your app. Its first session appears in Sypnia, and the first problem arrives in plain words.