Skip to content

v1.3.2

Choose a tag to compare

View all tags
@AkashAi7 AkashAi7 released this 23 Apr 11:05
· 3 commits to main since this release
v1.3.2
9b4fb1b

[1.3.2] — 2026-04-23

Fixed

  • Faster coordination flow: Coordination-only prompts such as task assignment, routing, and handoff requests now bypass the heavy tool-calling execution loop, which makes inter-agent assignment materially faster and avoids wasting turns on simple orchestration work.
  • Prompt bloat control: Continuation details, handoff packets, and attention-queue previews are now compacted so follow-up tasks stop recursively carrying oversized context blobs through the run.

Added

  • Exact model pinning: Added pixelSquad.copilotModelId and pixelSquad.claudeModelId so you can pin a specific model instead of relying only on provider-family heuristics.

Changed

  • Lower token overhead: Session logs, handoff summaries, inbox messages, and prompt serialization now clip and cap payload size to reduce token consumption during agent-to-agent calls.
  • Tighter execution bounds: Reduced the maximum tool-calling rounds from 12 to 8 so confused runs fail faster instead of burning extra time and tokens.

Full Changelog: v1.3.1...v1.3.2

Assets 4
Loading