Releases·Tool Release
OpenRod launches open-source control plane adding sandboxing and policy controls for Claude Code, Codex, Cursor agents
The OpenRod team launched an open-source control plane that sandboxes autonomous coding agents like Claude Code, Codex, and Cursor, building on NVIDIA's open-sourced OpenShell isolation layer. It adds one-click sandboxes, policy controls, and audit tooling, and the builders are soliciting community feedback on usability and agent support.
3 engagement1 sourceSat, Oct 3, 2026, 04:43 PM
AI Overview
- OpenRod, an open-source control plane built on NVIDIA's OpenShell, sandboxes Claude Code, Codex, and Cursor at launch
- One-click sandboxes run locally or on a user-owned remote host over SSH
- Policy controls include egress rules, groups, and secrets passed through a gateway
- Bring-your-own MCPs and Skills, plus an audit trail with SIEM export and webhooks
- OpenShell, open-sourced by NVIDIA, isolates agents at the file, network, process, and credential level
- Installable via npx openrod; repo at github.com/OpenRod/OpenRod
Related
Products·launched
Isolade: local-first coding agent workbench with secretless microVMs
Isolade is an open-source, local-first workbench that runs coding agents in isolated microVMs, combining agent management with microVM isolation. It provides each agent its own microVM with subsecond provisioning and domain-scoped secret substitution, so the VM only sees placeholder values, addressing the need for secure, isolated agent execution. It targets developers who want to run coding agents locally without exposing secrets.
Isolade is an open-source, local-first workbench that runs coding agents in isolated microVMs, combining agent management with microVM isolation. It provides each agent its own microVM with subsecond provisioning and domain-scoped secret substitution, so the VM only sees placeholder values, addressing the need for secure, isolated agent execution. It targets developers who want to run coding agents locally without exposing secrets.
1 engagement1 sourcehackernewsTue, Aug 4, 2026, 12:42 PM
Releases·Tool Release
NVIDIA launches Agent Toolkit with OpenShell runtime and AI-Q Blueprint
NVIDIA announced the Agent Toolkit, including the OpenShell open-source runtime for building self-evolving agents with improved safety and security. The toolkit also features the AI-Q Blueprint for agentic search, built with LangChain, which tops DeepResearch Bench accuracy leaderboards and can cut query costs in half.
NVIDIA announced the Agent Toolkit, including the OpenShell open-source runtime for building self-evolving agents with improved safety and security. The toolkit also features the AI-Q Blueprint for agentic search, built with LangChain, which tops DeepResearch Bench accuracy leaderboards and can cut query costs in half.
0 engagement1 sourcenvidianews.nvidia.comThu, Aug 6, 2026, 10:50 AM
Releases·Tool Release
h5i packages rootless container sandbox for Claude Code and Codex
A Medium post introduces h5i, a lightweight rootless container sandbox for running coding agents like Claude Code and Codex securely. It uses filesystem isolation, restricted networking, scoped credentials, and auditable execution to protect the host environment.
A Medium post introduces h5i, a lightweight rootless container sandbox for running coding agents like Claude Code and Codex securely. It uses filesystem isolation, restricted networking, scoped credentials, and auditable execution to protect the host environment.
4 engagement2 sourcesredditWed, Aug 12, 2026, 03:31 AM
Products·building
Tome: Security-first desktop cockpit for coding agents
Tome is a desktop app for macOS and Linux that integrates coding agents, terminals, editors, and an AI assistant into a tiling workspace with a sandbox around agents. It helps developers orchestrate agents, build flows, and learn new technologies securely. The app emphasizes security by sandboxing agents to prevent unauthorized actions.
Tome is a desktop app for macOS and Linux that integrates coding agents, terminals, editors, and an AI assistant into a tiling workspace with a sandbox around agents. It helps developers orchestrate agents, build flows, and learn new technologies securely. The app emphasizes security by sandboxing agents to prevent unauthorized actions.
4 engagement1 sourceredditTue, Aug 18, 2026, 08:57 PM
Products·launched
Watchfire: open-source control room for AI coding agents
Watchfire is an open-source control room that lets developers run multiple AI coding agents simultaneously without manual permission clicks. It evolved to support secure execution, spec-based development with tasks, and can be exposed as an MCP server. It solves the problem of managing and coordinating multiple agents efficiently.
Watchfire is an open-source control room that lets developers run multiple AI coding agents simultaneously without manual permission clicks. It evolved to support secure execution, spec-based development with tasks, and can be exposed as an MCP server. It solves the problem of managing and coordinating multiple agents efficiently.
1 engagement1 sourcehackernewsTue, Aug 4, 2026, 08:57 PM