Skip to content
| Marketplace
Sign in
Visual Studio Code>Other>Kiro AutoRunNew to Visual Studio Code? Get it now.
Kiro AutoRun

Kiro AutoRun

AN LE

|
280 installs
| (0) | Free
Auto-clicks Run/Accept All for Kiro IDE prompts with banned-keyword safety (macOS + Windows)
Installation
Launch VS Code Quick Open (Ctrl+P), paste the following command, and press enter.
Copied to clipboard
More Info

Kiro AutoRun

Visual Studio Marketplace GitHub Release Installs License

100% Native, zero-overhead auto-approval for Kiro IDE command prompts with banned-keyword safety. macOS, Windows, & Linux.

The Problem

Kiro IDE requires manual approval for commands the AI agent runs. This creates frustrating interruptions:

  • "Waiting on your input" panel appears continuously, breaking focus and flow.
  • "Accept All / Reject All" prompts block tasks endlessly.
  • Lost Sessions hang indefinitely awaiting action if you step away.
  • Built-in trustedCommands can be buggy, fails occasionally, and requires constant manual maintenance.

How It Works: Layer 0 Native Architecture

Kiro AutoRun has evolved from an OCR-based system into a 100% Native TypeScript Layer 0 API Integration.

Instead of reading the screen or taking over your mouse, it hooks directly into Kiro's internal execution state.

  • Zero Overhead: No Python, no heavy OCR, 0% CPU consumption.
  • Zero Interruption: Interacts instantly inside the VS Code backend—your cursor and window focus remain completely unaffected.
  • 100% Accurate: Checks internal command strings via Kiro IDE's kiroAgent.executions.getExecutions API natively, avoiding visual false positives.
  • Universal OS Support: Since it runs purely within the TypeScript extension host, it works out-of-the-box on macOS, Windows, Linux, and Remote Workspaces!

Key Features

  • Layer 0 Native APIs - Instantly reads internal pending executions and invokes executeCommand('kiroAgent.execution.runOrAcceptAll') directly.
  • Smart safety (Deny-list) - 28+ banned keywords are intercepted. Inherently dangerous commands (like rm -rf / or curl | bash) are blocked before they can run.
  • Auto-learning (Machine Memory) - Tracks how often a specific tool/command is invoked safely. Once a command proves benign repeatedly, it is dynamically learned and added to Kiro's internal trustedCommands.
  • NEVER_LEARN strict guard - Highly dangerous base commands (rm, sudo, curl) are systematically guarded and will never be auto-trusted.
  • Premium Ops Dashboard - A modern, glassmorphic UI integrated right inside Kiro to visualize your approval metrics, blocked commands, and up-time gracefully.

Installation

From Marketplace (Recommended)

Install from Visual Studio Marketplace

Download VSIX directly from Marketplace

Or search for "Kiro AutoRun" in the Extensions panel (Cmd+Shift+X).

From GitHub Releases

  1. Download the latest .vsix from GitHub Releases
  2. In Kiro/VS Code: Extensions > ... > Install from VSIX > select the downloaded file

Settings

Setting Default Description
kiroAutorun.enabled true Enable/disable Native auto-approval
kiroAutorun.pollInterval 2 Seconds between checking the internal execution queue
kiroAutorun.bannedKeywords (28 dangerous patterns) If any execute natively, auto-run is blocked

Legacy targetApp, OCR limits, and Recovery settings have been permanently retired due to the modern Native architecture.

Commands

Command Description
Kiro AutoRun: Toggle ON/OFF Toggle via command palette or status bar click
Kiro AutoRun: Show Command History Access the sleek Native API Dashboard log
Kiro AutoRun: Clear History Clear the Ops Dashboard history
Kiro AutoRun: Add Banned Keyword Add a keyword to the deny list
Kiro AutoRun: Add Trusted Command Pattern Add a pattern manually
Kiro AutoRun: Dump Available Commands Discovery tool to dump all internal API definitions

Safety Features

  • Banned keywords - intercepts native command streams prior to execution approval.
  • Pipe-to-shell detection - denies obscure background executions routing to untrusted bins.
  • NEVER_LEARN - strictly denies training behavior against OS-level destruction patterns. Default banned keywords include: rm -rf, sudo rm, chmod 777, curl | sh, git push --force, drop table, shutdown, kill -9, fork bombs, reverse shells, etc.

Author

Le Van An (Vietnam IT)

GitHub

Support

If you find this project useful and it saves you hours of clicking, consider supporting the author!

Bank Transfer

Method Account Name
MB Bank 0360126996868 LE VAN AN
Momo 0976896621 LE VAN AN

Shopee Affiliate

Shopee

View products on Shopee - Just one click helps! Thank you!

License

MIT License - Copyright (c) 2026 Le An (Vietnam IT)

  • Contact us
  • Jobs
  • Privacy
  • Manage cookies
  • Terms of use
  • Trademarks
  • Your Privacy Choices
  • Consumer Health Privacy
© 2026 Microsoft