[eric] release: cut 1.7.8-exp.13 with the workflow rescue, real-agent runs and the tucked browser miniature

This commit is contained in:
ciregenz
2026-08-17 14:47:33 -07:00
parent 4ac484bf3b
commit faecb120fa
2 changed files with 2 additions and 1 deletions
+1
View File
@@ -98,6 +98,7 @@ P_RELEASES: List[ReleaseNote] = [
"The mouse-wheel Zoom/Scroll setting works on real mice now. Accelerated wheels (Magic Mouse, Logitech smooth scrolling) report fractional scroll amounts that were being mistaken for a trackpad, so the wheel always panned no matter what the setting said.",
"A browser helper that talks itself into refusing (\"I\u2019m a text-based AI\") no longer poisons its browser for every later task. Refused and fabricated runs are forgotten instead of remembered, and agents can ask for a completely fresh browser when one misbehaves.",
"Agents no longer claim \"the user told me to stop\" when it was OpenSwarm\u2019s own housekeeping talking. Internal wrap-up and retry messages now identify themselves, so an agent\u2019s explanation of why it stopped reflects what actually happened.",
"Workflows you describe in chat can no longer save as empty shells that fail every scheduled run, pressing Run now opens the actual working agent instead of a separate monitor panel, and an agent's browser tucks neatly under its chat pill instead of floating beside it.",
"The amber resume chip now appears only on chats that were genuinely cut off mid-answer instead of every stopped chat, and agents can now see and operate the popup sign-in windows that sites open, instead of going blind at them.",
"Double-clicking empty canvas reliably fits every card on screen again, and an app an agent builds or drives now docks as a small card beside the chat instead of covering it. When a website needs you to sign in, the agent now pauses, tells you, and continues by itself once you log in on its browser card.",
"Agents no longer lose their tools partway through a run. Under heavy multi-agent load, a safety timer could mistake a queued browser or app task for a lost one and restart the agent's tool connection mid-task; it now checks the right task before acting.",
+1 -1
View File
@@ -1,6 +1,6 @@
{
"name": "openswarm",
"version": "1.7.8-exp.12",
"version": "1.7.8-exp.13",
"license": "AGPL-3.0-only",
"description": "OpenSwarm \u2014 AI Agent Orchestrator",
"author": "openswarm-ai",