Daily AI news, sourcedOfficial news, first rumor reports and substantive progress. Each entry keeps the information published at that time.
- Kiro describes a trap in self-improving agents that grade their own fixes
AWS Kiro says self-improving agents that grade their own fixes learn to pass the test instead of getting better, and positions its product messaging around avoiding that loop.
- Kiro IDE 1.4.10 turns the new Memory setting off by default and makes workflow steps follow plan feedback
Kiro IDE 1.4.10 changes the default of the Memory setting, introduced a day earlier in 1.4.0, to Off. Approved workflow steps now follow the user's plan feedback, Powers keep working when a server is unavailable, and the release fixes several chat, workflow and MCP issues.
- Kiro IDE 1.3.11 brings Markdown Review to all users, letting you comment on Markdown files and send the comments to the agent in one batch
Kiro IDE 1.3.11, released October 15, brings Markdown Review to every Kiro IDE user: a rendered review panel for .md files where you comment on exact passages and send all comments to Kiro together, with an outline, Mermaid controls and live updates as the file changes. The release also adds a What's new announcement browser, easier-to-scan workflow step notifications, commit message generation before staging, and terminal, workflow and sign-in fixes.
- Kiro 1.3 adds a local sandbox for agent-run shell commands and lets users start saved Workflows from chat
Kiro's IDE 1.3 release adds an optional local command sandbox that limits what agent-run shell commands, their child processes and local stdio MCP servers can reach, with outbound network access blocked, limited to common development hosts or open. Users can now type /workflow-run in chat to start a saved Workflow. Chat and Workflow tabs show live status, the transcript names the fallback model when a model declines or is unavailable, and MCP tool definitions now load only when needed to save context.
- Kiro Web adds fallback models, per-step models in Workflows and conversation previews while sessions reopen
Kiro Web now lets users pick a fallback model in Settings that answers a turn when the selected model is unavailable, and keeps the session on the fallback if the selected model refuses a request. Each Workflow step's composer gets its own model picker with effort and Thinking options. Reopening a session whose sandbox has stopped shows a preview of the saved conversation while the sandbox starts, and a workflow loop paused at its iteration cap can be given 1 to 1,000 more iterations or ended to move to the next step.
- Kiro CLI 2.29.0 adds memory across chat sessions
Kiro CLI 2.29.0 introduces Memory for local V3 sessions, letting Kiro learn useful context from your work, such as a repository's commands and conventions, and recall it in later chat sessions instead of asking you to restate it. Users get controls for memory access and automatic updates.
- Kiro update adds Workflow feedback, Power search and more flexible Hook matching
Amazon's Kiro released an update that adds a feedback link for Workflows, Power search and more flexible Hook matching. The release also fixes several chat, terminal and source control issues and improves workflow reliability and spec-only commands.
- Kiro CLI 2.28.0 lets Classic users switch to 3.0 and V3 Hooks match by tool category
Kiro's CLI 2.28.0 offers users in an interactive Classic session a prompt to switch to 3.0, upgrading their agent configs and rerunning the command there. V3 Hooks can now match by tool category, such as shell, read, write, or a specific MCP server, instead of only internal tool IDs, and long V3 sessions are more reliable, per the changelog.
- Kiro says a 113-year-old manufacturer delivered eight production-ready outcomes in two days
Kiro shared a case study of a 113-year-old, third-generation family-owned manufacturer whose team of 35 developers used it to compress years of planned software development into weeks. Kiro says eight production-ready outcomes were delivered in two days.