v1.33.7
CodeSplash AI 1.33.7
This is a focused Command Center patch release aimed at making AI agent launches more dependable. If you use Command Center to spin up parallel agents across repos, 1.33.7 should give you a smoother launcher flow, clearer queue behavior, and better feedback when runs exit.
Fixes
- Fixed AI agent launcher issues in Command Center to improve launch reliability.
- Simplified shared launch behavior so the launcher is less likely to rely on brittle CLI-specific handling.
- Removed ambiguous per-run-mode launch logic that could lead to unsupported flags and fast-failing agent sessions.
- Improved how exited sessions are handled and displayed in the Command Center grid.
Improvements
- Refined the launch queue experience with clearer row/placeholder behavior during multi-agent launches.
- Tightened opening prompt delivery across agent CLIs to make starts more consistent.
- Expanded exit code handling so Command Center can provide better feedback when an agent stops.
- Standardized agent names across the UI for a cleaner launcher experience, including labels like Codex, Gemini, OpenCode, ForgeCode, Kiro, and Copilot.
- Removed attachment-related launch flow complexity from Command Center as part of this stabilization pass.
Notes
- This release is primarily a reliability and UX polish update for Command Center rather than a major feature drop.
- If Command Center is part of your daily multi-agent workflow, this is a recommended update.