Improvement April 2, 2026 v0.4.2
Faster context synthesis
Serv now queries GitHub, Sentry, and your flag provider in parallel rather than sequentially. Average time from @serv mention to first response dropped from 6.4s to 2.1s.
- + Parallel integration queries for all context requests
- + Sentry stack trace truncation now respects token limits more gracefully
- + Fixed an edge case where PagerDuty escalation fired twice on retry
Fix March 24, 2026 v0.4.1
Incident channel invite reliability
A regression in 0.4.0 caused responder invites to fail silently when the on-call schedule returned more than 3 names. Fixed and covered with integration tests.
- + Fixed silent failure on PagerDuty schedules with large rotation counts
- + Mattermost slash command responses now include the incident ID consistently
- + Improved error messaging when a connected integration token has expired
Release March 17, 2026 v0.4.0
GrowthBook, Alertmanager, and Datadog support
Three new integrations land in this release: GrowthBook for teams running open-source feature flagging, Alertmanager for Prometheus-native alert routing, and Datadog monitor webhooks for inbound incident declaration.
- + GrowthBook: read flag state, toggle experiments, adjust rollout percentages
- + Alertmanager: inbound webhook support with label and annotation parsing
- + Datadog: inbound monitor webhook with priority and tag context
- + Overage usage bar now visible in workspace settings
- + AI query count shown per incident in the channel header
Improvement March 5, 2026 v0.3.2
Postmortem drafting improvements
The postmortem draft Serv generates at incident close is now structured around the timeline of actions taken in the channel, not just the final state. Includes a contributing factors section and suggested follow-up tasks.
- + Postmortem now includes full action timeline from channel history
- + Contributing factors section derived from flag changes and deploy diffs
- + Suggested follow-up tasks based on the incident type and affected services
- + Postmortems can now be exported as Markdown or sent to a configured Notion workspace
Fix February 19, 2026 v0.3.1
GitLab MR revert reliability
Revert MRs created by Serv on GitLab occasionally targeted the wrong base branch when the default branch was not named main. Fixed by reading the project default branch from the API before creating the revert.
- + Fixed GitLab revert MR targeting wrong base branch
- + GitHub revert PR now includes the incident channel link in the PR description
- + Sentry query now falls back to project-level stats when org-level access is restricted
Release February 10, 2026 v0.3.0
GitLab and Opsgenie support
GitLab is now a first-class code integration alongside GitHub. Serv can read MR history, diff responsible merges, and create revert MRs without leaving the incident channel. Opsgenie joins PagerDuty as a supported paging provider.
- + GitLab: read commits, diff MRs, create revert MRs and hotfix branches
- + Opsgenie: on-call queries, alert creation, escalation chain support
- + Slack: improved thread handling for incidents in busy channels
- + Status page updates now include an estimated resolution time field
Release January 22, 2026 v0.2.0
Built-in status page and knowledge base
Serv now ships with an integrated status page that can be updated from inside the incident channel. No third-party dependency required. The org knowledge base lets teams store runbooks and past incident summaries that Serv references when responding.
- + Built-in status page with per-service component status
- + Status updates posted from the incident channel in one command
- + Org knowledge base for runbooks, playbooks, and past incident summaries
- + Serv surfaces relevant runbooks automatically when an incident opens
- + LaunchDarkly: flag change history now shows the user who made the change
Release January 6, 2026 v0.1.0
Early access launch
Serv opens to early access. Core incident flow is live: automated channel creation, PagerDuty paging, GitHub and Sentry context, LaunchDarkly flag management, and an AI co-responder available via @serv in any incident channel.
- + Automated incident channel creation in Mattermost and Slack
- + PagerDuty on-call queries and incident declaration
- + GitHub commit history, PR diff, and revert PR creation
- + Sentry error rate, stack trace, and impact queries
- + LaunchDarkly flag state, toggle, and rollout adjustment
- + AI co-responder available via @serv in incident channels