nolabs-aiDeveloper tools
nono
A zero-latency, zero-setup sandbox for AI agents that enforces least-privilege isolation for both the agent and its delegated tools.
- Backend
- DevOps
- Security
- Automation
- CLI
- Library/framework
- macOS
- Linux
- Windows
- Runs locally
- Docker supported
- Popularity
- 3.8k Stars
- GitHub stars
- Recent activity
- 8/24/2026
- Updated in the last 30 days
- License
- APACHE-2.0
- Permissive
Why it matters
We look beyond stars: what problem it solves, whether it creates real utility, and what makes its approach worth noticing.
Problem
Addresses the security isolation challenge for local AI agents, preventing agents and their invoked tools (e.g., git, kubectl) from unauthorized access to sensitive files, networks, or credentials.
Practical value
Provides a zero-setup sandbox environment for macOS, Linux, and WSL2, enabling granular control over filesystem, network, and credentials via JSON configuration, including nested sandbox isolation for toolchains.
Innovation / differentiation
Unlike traditional container or VM approaches, nono utilizes a lightweight sandbox mechanism and introduces independent policy brokers for child processes (tools), enabling credential injection and permission restrictions based on L7 filtering.
Leverage potential
Offers FFI bindings for Rust, Python, TypeScript, and Go, and supports sharing/extending profiles via a registry, facilitating the standardization of agent security policies in team and production workflows.
Why now
As AI agents become more integrated into development workflows, the demand for security and auditability in local execution environments is increasing, positioning this tool to address a current gap in agent security governance.
Community activity
In the last 90 days there were 293 new issues and 389 pull requests; the bounded issue/PR samples include 34 issue authors and 19 PR contributors, with at least 20 releases.
Maintainer responsiveness
The 100-issue window sample had a 86% close rate, and the 100-pull-request sample had a 83% merge rate. Maintainer-response observations covered 66% of that issue sample, with a 0% response rate and median first response of not enough data.
Key highlights
- Provides a zero-latency, zero-setup sandbox environment for running AI agents
- Enforces least-privilege access to filesystems, networks, and credentials
- Supports isolated child sandboxes for individual tools invoked by agents
Quick start
How it is installed, how hard it is, and where to start.
Where it runs
Runs locally
Difficulty
Easy — follow the steps
- 01Install the nono CLI tool using curl.
- 02Search for an agent profile using the nono search command.
- 03Run the agent within a specific profile using the nono run command.
Best for
- Developers who want to try it on their machine
- People who prefer Docker deploys
More about it
nono is a secure execution environment designed for AI agents. It bypasses traditional container or VM overhead by using process-level isolation, ensuring agents are restricted to authorized filesystem paths and network resources.
Beyond agent isolation, nono applies independent sub-sandbox policies to tools invoked by the agent, such as git, gh, or kubectl. This prevents privilege escalation, ensuring that even if an agent is compromised, its delegated tools remain restricted.
Sources
Each field shows its status and source — expand to review.
12 · Expand
Sources
Each field shows its status and source — expand to review.
capability tags
Verifiedautomation, security
Source: admin_cms · cms editor · 8/18/2026
Latest release
Verifiedv0.74.0
Source: GitHub API · latest_release=v0.74.0 · 8/25/2026
License
VerifiedApache-2.0
Source: GitHub API · license.spdx_id=Apache-2.0 · 8/25/2026
needs api key
VerifiedNo
Source: admin_cms · cms editor · 8/13/2026
One-liner
Verified{"en":"A zero-latency, zero-setup sandbox for AI agents that enforces least-privilege isolation for both the agent and its delegated tools.","zh":"为 AI Agent 提供零延迟、零配置的沙箱运行环境，通过最小权限原则隔离代理及其调用的工具。"}
Source: heuristic_batch_publish · batch-publish-tools heuristic write · 8/24/2026
Platforms
Verifiedmacos, linux, windows
Source: admin_cms · cms editor · 8/13/2026
Category hint
Inferred from materialsai-apps
Source: Project README · hint=ai-apps · 8/13/2026
product forms
Verifiedcli, library_framework
Source: admin_cms · cms editor · 8/18/2026
role tags
Verifiedbackend, devops, security
Source: admin_cms · cms editor · 8/18/2026
supports docker
VerifiedYes
Source: Repository file · dockerfile=true; compose=false · 8/25/2026
supports local
VerifiedYes
Source: admin_cms · cms editor · 8/13/2026
supports self host
VerifiedNo
Source: admin_cms · cms editor · 8/13/2026
Related projects
Other verified projects matched by category, capabilities, and intended roles.
microsandbox
A fast, local-first microVM runtime and library for running untrusted workloads.
open-connector
An open-source agent connector gateway that keeps third-party app accounts behind a dedicated runtime and exposes actions through web, MCP, HTTP/OpenAPI, and SDK interfaces.
mcptoon
A lightweight MCP CLI client that reduces AI agent context usage by compressing tool manifests and execution results.
container
Apple’s command-line tool for running OCI Linux containers as lightweight virtual machines on Apple silicon Macs.
zx
Google’s open-source JavaScript/TypeScript scripting tool with convenient cross-platform wrappers for shell command execution.
compose
Docker’s official CLI tool for defining and running multi-container applications from a Compose file.