Short answers first. Where the answer is "no", it says no.
SnagSpy is a hosted error monitoring and observability platform. It ingests errors, logs, traces, metrics and session replay over standard OpenTelemetry, correlates them, and attributes a failure to the change that caused it. It is aimed at engineering teams that want an explanation of an incident rather than a larger pile of signals.
Three differences matter. Root-cause findings are graded High, Medium or Low against actual evidence rather than scored with an invented percentage. Deployments and configuration changes feed the analysis instead of sitting beside it. AI investigations are included in the plan allowance rather than billed per contributor as an add-on. Sentry is better at profiling, fix generation and ecosystem breadth.
Yes, and it is the primary way data arrives. Errors, logs, traces and metrics are accepted over OTLP at standard endpoints. An existing OpenTelemetry setup is repointed rather than rewritten, and the same instrumentation keeps working if you later move to another vendor. Instrumentation is not proprietary.
There is a free plan at no cost, a Team plan and a Business plan billed monthly in US dollars, and a negotiated Enterprise plan. The paid tiers differ mainly in retention, session replay and AI investigations rather than in event volume. Current figures are on the pricing page, which is generated from the same catalogue the platform bills from.
Usage past an allowance is reported to you, not silently discarded. Refusing a customer’s telemetry is the most destructive thing an observability platform can do, so it is never a surprise: the dashboard shows what was used against what was included, and the behaviour on each signal is published rather than discovered.
Retention is set by plan: seven days on the free plan, thirty days on Team, and ninety days on Business and Enterprise. A plan may only narrow the platform-wide retention ceiling, never widen it, and that rule is enforced by a test rather than by review.
No. SnagSpy is a hosted service and there is no self-hosted or on-premise licence today. If running the platform inside your own infrastructure is a requirement, SnagSpy is not currently the right choice and it is better to know that now.
It means a finding carries High, Medium or Low confidence, and shows what that grade rests on — a deployment that preceded the failure, a configuration change, independent signals that agree. SnagSpy deliberately does not print a figure like "98% confident", because no such number can be justified and invented precision invites teams to act on a measurement that was never taken.
Anything that can emit OpenTelemetry, which covers most modern server and browser runtimes, plus a JavaScript SDK published from this project for browser error capture and session replay. SDK coverage is narrower than long-established competitors; if you need a first-party SDK for an unusual runtime, check before committing.
Yes. Scrubbing runs in the SDK before data leaves your application, so common sensitive fields are removed at source rather than after arrival. Session replay is gated by per-project consent. Retention ceilings and the wider security posture are set out on the security page.
No. Neither UI profiling nor continuous profiling is built, and neither is on a near-term roadmap. If profiling is central to how your team debugs performance, that is a real gap and Sentry or a dedicated profiler is the better answer.