Kiro IDE: Local Command Sandbox, Start Saved Workflows from Chat, and Live Tab Status

This release lets you limit what agent-run shell commands can reach and start saved Workflows from chat. It also shows live session status on editor tabs, explains model fallbacks, loads MCP tools on demand, and fixes Workflow, chat, and account issues.

[Contain agent-run shell commands in a sandbox] Turn on the local command sandbox to limit what agent-run shell commands and their child processes can access. Choose whether outbound network access is blocked, limited to common development hosts, or open. You can also run local stdio MCP servers inside the same sandbox.

[Start saved Workflows from chat] Type `/workflow-run`, choose a recipe, complete its inputs, and start the run. Kiro opens Workflows on the new run. Learn more ->

[Improvements] On-demand MCP tools: Kiro now loads MCP tool definitions only when needed, leaving more of the model context available for your work. Learn more

[Improvements] Live status on editor tabs: Chat and Workflow step tabs now show when Kiro is working or needs your input, then return to their normal icon when the session goes idle. Learn more

[Improvements] Model fallback notices: When a model declines or is unavailable, the transcript now names the model that continues and whether it applies to one turn or the rest of the conversation. Learn more

[Improvements] Clearer empty views: Empty Workflows and Artifacts views now explain what belongs there and how to add the first item

[Fixes] Active Workflow runs now stay above finished runs, and rows no longer jump when a run pauses or needs input

[Fixes] Workflow age labels now update while the list is open, and expandable rows show a chevron

[Fixes] A single active Workflow now opens expanded, stopped runs no longer show as waiting, and parallel steps no longer appear twice

[Fixes] Profile, usage, and plan details no longer wait indefinitely when requests stall

[Fixes] Text in animated WebP images now stays sharp when Kiro reads the first frame

[Fixes] MCP tools on streamable HTTP servers keep working after a server drops its session

[Fixes] Two- and three-digit list numbers now remain fully visible in the chat input, agent replies, and tool cards

[Fixes] Workflow recipe tables now fit narrow Agent Focus windows without pushing controls outside the panel

[Contain agent-run shell commands in a sandbox] The agent, built-in file tools, and remote MCP servers stay outside the sandbox. Permission rules still govern agent actions. The sandbox reduces risk but cannot guarantee complete isolation, so keep reviewing proposed commands. Select Command sandbox in the status bar to open the settings and confirm the sandbox is active. Learn more ->