# Command Bridge launches

**Date:** 2026-04-13 | **Period:** Week of April 13, 2026 | **Edition:** weekly | **Canonical URL:** https://outcomet.app/changelog/2026-04-13-command-bridge-launches

A big platform week, agents, integrations, and AI safety all moved forward. Nova
steps up as a primary Crew agent under explicit approval flows, GitHub and Jira
sync both ways, and themes can be exported straight to issue trackers.
Highlight: **Command Bridge launches** as a single home for agents,
capabilities, and team rosters.

---

## Command Bridge launches

Command Bridge is now live, a single interface that pulls together your agents,
active learning clusters, and team rosters so you can see who (and what) is
working on what without hopping between screens. It pairs with a new Active
Roster side panel in CrewOverview and redesigned crew cards (portrait split
layout, larger hero visuals) so live capability state is readable at a glance.

- New Command Bridge: agents, learning clusters, and rosters in one view
- Active Roster side panel in CrewOverview
- Redesigned crew cards: portrait split layout + larger hero visuals
- Faster path to find the right person or agent capability

---

## Agents, Nova & AI safety

Nova steps up as a primary Crew agent this week, with her own avatar and visual
presence, and the whole AI surface gets a safety pass. Background actions that
mutate data now require explicit approval, admins get a workspace-level opt-in
for AI & Automation, and you can pick the model (Haiku, Sonnet, Opus) that
matches the tone and depth you want. Background tool activity and AI job results
now show up in the chat timeline so nothing happens behind your back.

- Nova promoted to primary Crew agent with avatar and visual presence
- Explicit approval flow for mutating actions, admin-controlled AI & Automation
  opt-in
- Read-only fallback when mutations are disabled
- Agent model selector: Haiku, Sonnet, Opus
- Background tool activity and AI job results visible in the chat timeline

---

## Conversations & memory

Long sessions hold together better. Context compaction and sliding-window
summarization keep the relevant pieces in scope without losing key details, long
messages no longer get truncated or fail silently, and the agent timeline
handles unusual step types more gracefully. Fetching is also lighter, so
responses feel snappier.

- Improved long-conversation memory (compaction + sliding window)
- Long messages handled correctly, agent timeline more resilient
- Lighter client/server payloads for snappier responses

---

## GitHub & Jira: two-way sync, plus theme export

Issue trackers stop drifting from your workspace. GitHub and Jira issue
descriptions now sync **both ways**, pull external updates in, push local edits
back, and Jira specifically gets better formatting interpretation and
closed-issue handling. From the Changes dashboard, GitHub PRs open directly in
their original tab. The bigger move: you can now export themes straight to
GitHub Issues or Jira Epics/Stories with configurable field mapping and
automatic link tracking, and push a single theme to multiple connections at
once.

- Bi-directional GitHub & Jira description sync
- Better Jira formatting and closed-issue handling
- One-click open of source PRs from the Changes dashboard
- Theme export to GitHub Issues and Jira Epics/Stories with field mapping
- Multi-connection push for a single theme

---

## Feedback & insights

Feedback gets sharper classification, a redesigned detail view (two-column
desktop with a Nova Insights collaboration sidebar), and **anonymous
submission** so candid input is easier to collect. Severity (S0-S3) was refined
so impact reads cleanly across the triage flow.

- More accurate feedback classification and analysis
- Anonymous feedback submission supported
- Redesigned FeedbackDetail with Nova Insights sidebar
- Refined S0-S3 severity indicators for clearer prioritization

---

## Admin, onboarding & connection management

Org admins can now set a **Default Role** (Viewer, Member, Admin) for new users
right from the Members interface. Setup got smoother too: conditional discovery
tabs, simplified rule management, and a unified smart search/filter bar across
the connection catalog and active connections, including better behavior on
smaller screens.

- Default Role for new users (Viewer/Member/Admin) in Members
- Conditional discovery tabs and simpler setup/rule management
- Unified search + filter across connection catalog and active connections
- Better responsiveness on smaller screens

---

## UI, reliability & background jobs

A round of polish across discovery and setup flows, a unified EmptyState for
Discovery/Signals/Capabilities/Bridge, and a fix for the Capabilities Map so
dragging or misclicks outside grouped nodes no longer hijack your selection.
Underneath, the background job pipeline and AI Jobs Engine were upgraded,
results persist reliably and surface faster, and real-time subscriptions are
more stable, with broader error handling reducing silent failures across chat
and integrations.

- Unified EmptyState across Discovery, Signals, Capabilities, Bridge
- Capabilities Map: dragging/misclicks outside groups no longer hijack selection
- UI alignment polish across discovery and setup
- Upgraded background job pipeline + AI Jobs Engine, faster, more reliable
- More stable real-time subscriptions, broader error handling across chat &
  integrations

---

## What this means for you this week

- Command Bridge gives you one place to see your agents, capabilities, and team,
  fewer screens, faster context.
- Nova feels more present (avatar, primary Crew slot) and works under explicit
  approvals you can see in the chat timeline.
- Issue trackers stay honest, GitHub and Jira sync both ways, and you can push
  themes straight into them.
- Feedback collection gets candid input via anonymous submission, with cleaner
  detail views and Nova Insights alongside.
- Long agent conversations hold their context, and background AI work is faster
  and more reliable.

If you want a walkthrough of Command Bridge, theme export mapping, or the new
admin/AI safety controls, reply here and I'll schedule a quick demo.
