Open source · the affordable Sentry alternative →

Capture every error.
Agent-ready fixes.

Start by capturing production exceptions in real time. Then connect your AI agent via MCP — it reads the stack traces, investigates the root cause, and ships the fix autonomously.

Limited early access · free for founding teams · no credit card

app.trycatch.dev — Issues · your-org
{} trycatch
A your-org
Issues 24
Dashboard
Releases
Alerts
Configuration
Projects
Members
your-org / Issues
Total errors
1.4k
↑ 12% vs yesterday
Users affected
38
↑ 3 new
Resolved 24h
17
↓ improving
P50 latency
89ms
↓ 4% better
Unresolved
Level: Error
Project: All
● fatal
TypeError: Cannot read properties of undefined (reading 'price')
checkout.js:142 · PaymentForm.submit()
2,841
3m ago
error
● error
NetworkError: Failed to fetch /api/orders/list
api.js:88 · OrdersService.list()
917
12m ago
error
● warning
ReferenceError: cartItems is not defined
cart.js:56 · CartContext.reducer()
234
1h ago
warn
● error
ChunkLoadError: Loading chunk 42 failed
webpack.runtime.js:12 · LazyComponent()
87
3h ago
info
A better alternative to Sentry
Open source Self-host for free. Audit the code. Contribute. No black boxes.
Radically affordable 100k errors/month free. Team plan from $29. No per-seat pricing.
Agent-native MCP-first API so your AI agent can debug and fix errors autonomously.
Features

Everything you need to debug with confidence

Built by engineers, for engineers — no bloated dashboards, just the signal that matters.

Real-time error tracking
Catch and deduplicate errors the moment they happen. See full stack traces, breadcrumbs, and exactly which users are affected.
Smart alerts
Alert rules that learn your baseline. Get paged when error rates actually spike — not on every exception. Slack, PagerDuty, webhooks.
Release health
See exactly which deploy introduced a regression. Flag releases with error spikes automatically and roll back with confidence.
Source maps & stack traces
Upload source maps and get human-readable stack traces — not minified gibberish. Inline code context shows exactly what broke.
Trend analytics
Track error volumes over time, correlate spikes to deploys, and measure how reliability improves sprint over sprint.
Team workflows
Assign issues, comment, set priorities. Integrates with GitHub, Jira, and Linear to turn errors into tracked tickets in one click.
Drop-in Sentry replacement

Already using Sentry?
Just swap the DSN.

trycatch is fully compatible with the official Sentry SDK. No code changes, no new dependencies — point your existing dsn to trycatch and you're done.

  • 1 Keep your existing Sentry SDK — no package changes
  • 2 Replace the dsn value with your trycatch DSN
  • 3 Errors start flowing in — no other changes needed
JavaScript
Python
Go
// No package changes needed — use @sentry/browser as-is
import * as Sentry from "@sentry/browser";

Sentry.init({
  // ✦ only change this line
  dsn: "https://key@ingest.trycatch.dev/42",
  release: "my-app@2.1.0",
  environment: "production",
});

// All Sentry APIs work unchanged — captureException,
// setUser, addBreadcrumb, withScope, and more.
Testimonials

Loved by engineering teams everywhere

★★★★★

"We cut our mean time to resolution by 60% in the first month. The source map integration is flawless — we finally see real stack traces in production."

SM
Sarah M.
Engineering Lead · Fintech startup
★★★★★

"Alert deduplication alone saved my team hours of on-call toil every week. We're not drowning in noise anymore — just the bugs that actually matter."

JC
James C.
Staff Engineer · E-commerce platform
★★★★★

"Switching took 20 minutes. Release health tracking caught a regression during deploy that would have hit 40% of our users before we noticed."

AL
Anaya L.
CTO · SaaS company
Pricing

Simple, transparent pricing

Start free. Scale when you're ready. No per-seat pricing, no surprise bills.

Developer
$0
forever free
  • 100,000 errors / month
  • 1 project
  • 7-day data retention
  • Email alerts
Most popular
Team
TBD
pricing coming soon
  • 1M errors / month
  • Unlimited projects
  • 90-day data retention
  • Slack + PagerDuty alerts
  • Release health tracking
  • GitHub / Jira integration
Enterprise
Custom
volume pricing · talk to us
  • Unlimited events
  • SSO / SAML
  • 1-year data retention
  • Dedicated SLA
  • On-premise option
  • Priority support
FAQ

Common questions

trycatch exposes an MCP (Model Context Protocol) server your agent can connect to. Once connected, the agent can query live errors, read full stack traces with source context, search by release or environment, and push resolution status back — all through standard tool calls. No custom glue code needed.
Any agent that supports MCP works out of the box — Claude, Cursor, Windsurf, and custom agents built on LangChain, LlamaIndex, or the OpenAI Agents SDK. We also provide a REST API and webhooks for agents that prefer direct HTTP calls.
Our SDKs cover JavaScript / TypeScript, Python, Go, Ruby, Java, and Rust. For other runtimes, the REST ingest API accepts any JSON payload. Source map upload is supported for all compiled-to-JS targets including TypeScript, esbuild, and webpack.
Yes. Error data is encrypted at rest and in transit, scoped strictly to your organization, and never used for training or shared with third parties. Enterprise plans include data residency options and a signed DPA.
Self-hosted deployment is available on the Enterprise plan. It ships as a single Docker Compose stack with Postgres and object storage — no Kubernetes required. Cloud-hosted and self-hosted instances share the same feature set.

Be first to catch every error.

Join the waitlist for early access. Founding teams get extended free quotas and direct input on the roadmap.

{}

Get early access

Founding teams get extended free quotas, direct roadmap input, and first access to AI-native error resolution.

No spam, ever · unsubscribe anytime