CodeRunner
Local sandbox environment for executing and testing AI agent code safely.
0 members list CodeRunner, unchanged over the last 12 weeks, data as of 3 November 2026
Members who list CodeRunner in their Stack, by week.
Be the first to add CodeRunner
Nobody on Stackness runs CodeRunner yet. Add it to your stack and you're the first member running it.
Homebrew analytics installs: 17 installs (2026) to 26 installs (2026)
Imported from Homebrew analytics
About CodeRunner
CodeRunner is a local sandbox environment for running and testing AI agent code safely. It provides isolated execution environments using hardware-isolated microVMs that can boot in under 200 milliseconds, allowing developers to execute untrusted code with controlled network access and runtime observability. Each agent gets its own Linux environment with full filesystem, networking, and package management capabilities. The tool supports various deployment patterns, from one-shot tasks to persistent long-running sessions, and includes features like snapshots, egress controls, and secret injection via proxy to prevent prompt injection attacks.
Developed by instavm, CodeRunner is written in Python and licensed under Apache 2.0. The platform integrates with major AI frameworks and offers both free and paid tiers, with a RESTful API and CLI tools available for Python and TypeScript. It includes full execution logging, network traces, and runtime event auditing for debugging purposes.
Checked 3 November 2026.
Nobody on Stackness runs CodeRunner yet.
Sign up and add it