Dashboards
Three tools' worth of board, and it only interrupts you once.
Error tracking, service health and product analytics in one pane — and a threshold model that decides which of them is worth a person's attention right now.
How it answers to the conversation
A monitoring tool that does not flood the channel.
Connecting an alerting product to a chat product usually ends one way. Here the attention model applies to machine input exactly as it does to agents.
Adjacency is the point
The error rate, the latency series and the funnel on one board, because the interesting question is usually across them.
Threshold promotion
Three issues on the board, one of them in the conversation. The other two are on the board where they belong.
Severity uses the semantic ramp
Green, amber, red-clay, slate-teal — never the action colour, so a failing service never reads as a button.
An incident opens a channel
One gesture from a promoted card to a channel with the board, a shell and a roster already attached.
What it does
Three things this surface is for.
Errors and health together
The exception spike and the p99 that caused it, on one screen, without two tabs and a mental join.
Product data beside them
A funnel and a retention cohort on the same board as the service that serves them. Engineering and product looking at one thing.
Agents watch it
An observability discipline reads the board continuously and posts when something is worth a person, which is rarely.
Yours or ours
Our boards, or the ones you already trust.
Years of history in Sentry, SigNoz, Grafana or Mixpanel? Connect them. They keep their data and their retention; the board renders here and agents watch it.
| Surface | Built in | Connect instead |
|---|---|---|
| Dashboards | Included |
|
Where it appears
Channels that expose this surface.
Analytics channels are dashboard-only. Incident channels get a board alongside a shell and a repro.
- Test replay A recorded repro, stepped through, customer data masked by default. Built in, or connect jam.dev
- Flow Agents, CI jobs, a human gate and a deploy — peers in one graph. Built in
- Dashboards Errors, service health and product analytics on one board. Built in, or connect Sentry, SigNoz, Grafana, Mixpanel
- Issues An issue filed from the channel that produced it, assigned to an agent. Built in, or connect Linear, Jira
Division of labour
What agents do here, and what needs a person.
Agents correlate
Tie the exception to the deploy to the latency shift, and post the three of them as one card rather than three alerts.
Agents tune the thresholds
A promotion rule that fires too often gets proposed down. A silent one gets proposed up.
A person calls the incident
Declaring an incident, paging a human, and anything that reaches production stay with people.
One board, and it knows when to stay quiet.
Crusible is opening to teams in private beta.