how we fit together

Wystan + Sentry

Sentry owns the error — the stack trace, the release, the user it hit — and with Seer it now reviews your pull requests with production telemetry in hand. Wystan sits one level up: it groups Sentry’s errors by the product feature they land in, joins them to usage, ownership and history, and tells every PR what it is touching. Sentry says this stack trace broke; Wystan says Checkout broke. We consume Sentry, we don’t replace it.

What Sentry does best

The system of record for production errors. Symbolicated stack traces, release health, session replay, profiling — and Seer, the AI agent that finds root causes and reviews PRs with production context, flat-priced per active contributor.

Error & performance monitoring with symbolicated stack traces
Release health, session replay, profiling
Seer: AI root-cause analysis + production-aware PR review
Flat $40 / active contributor / mo for Seer, unlimited usage
First-party MCP server

What Wystan adds

The product grain. Sentry attributes an error to an issue, a file and a release; Wystan attributes it to a named product feature and a user journey — and puts usage, ownership, AI share and history on the same row.

Errors grouped by product feature (“Checkout broke”), not by stack trace
Usage, ownership, AI share and history joined to the same feature
PR comment: which features the diff touches + their live error count
Weekly Slack digest: which features regressed this week
Per-org pricing, unlimited developers — zero SDK changes, we read your Sentry

using both

How they actually combine

Developer pushes a PR with changes to chargeCustomer() in the billing module.

Seer reviews the diff with production context — flags a null path it has seen fail in prod, suggests a guard.

Wystan layer adds “this PR touches Checkout (health 41, 9 Sentry errors in the last 7 days, 18% of traffic) and shared code in Subscriptions — blast radius 2 features; the author of the hottest file left in March.”

or alone

Just Wystan, on its own

Without Sentry, Wystan still maps, scores and comments — you lose the production error overlay, and keep everything that comes from git history: features, journeys, health, ownership, rework, AI share, blast radius. Without Wystan, Sentry tells you what broke and where in the code — but not which product capability it belongs to, how much traffic that capability carries, or how it has been trending. The two are designed to stack.

side by side

Where each one focuses

Focus area	Sentry	Wystan
Primary unit	Issue / stack trace / release	Product feature + user flow
Production errors	✓ the source	✓ grouped per feature (read from Sentry)
PR review	Seer — the diff, with production context	Feature map + risk + blast radius + ownership
History & trends	Release health	Health, rework, ownership per feature, over time
Usage	—	PostHog traffic per feature (risk × usage)
MCP server	✓	✓ + 13 tools
Pricing	Seer $40 / active contributor / mo	$99–299 / org / mo, unlimited developers
Best fit	Debugging and incident triage	Knowing what changed, what it touched, and what is getting worse

honest take

No vendor pressure

Sentry is the system of record for production errors, and Seer is the most direct overlap with what we do at PR time. We don’t compete on stack traces. Our slice is the product grain: Sentry says this trace broke; Wystan says Checkout broke, it carries 18% of traffic, its owner left, and this PR touches it. If your pain is debugging → Sentry. If it is knowing what your agents and humans changed, in product terms, and whether it is getting worse → Wystan. Most of our users run Sentry underneath.

also comparing

Other tools we work alongside