openagents releases launcher v1.1.0 with sign-in and model fixes

## What's Changed * Bugfix/cursor signin status by @QuanCheng-QC in * feat(workspace) offer signed-in Claude and Codex agents the models their own CLI lists by @QuanCheng-QC in * fix(opencode) finish an install whose postinstall was skipped instead of recording a placeholder as installed by @QuanCheng-QC in * Keep Stop working after a workspace response is cut short by @QuanCheng-QC in * fix(auth) detect an Antigravity account sign-in on Windows and clear the dead ends around it by @QuanCheng-QC in * fix(chat) keep the waiting bubble when switching threads by @QuanCheng-QC in * fix(workspace): let Anthropic-compatible endpoints verify and pick models; launcher 1.0.18 by @Ashin-LX-98 in * feat(workspace): renewable sessions, Apple sessions, and deleted workspaces refuse their token by @lihongmeng in * fix(hermes) keep one session per thread and send the instructions as a prefill by @QuanCheng-QC in * fix: keep non-gRPC transports optional by @simonyang08 in * fix(workspace): route cloud agent DM replies to the sender by @lgchao28-ops in * Release Workspace v1.1 with backward-compatible ownership by @zomux in * Preserve duplicated authorization headers for legacy agents by @zomux in * perf(workspace): reduce polling and file permission query overhead by @zomux in * perf(workspace): streamline middleware and reuse resolved workspaces by @zomux in * fix(workspace): k