eugene1g/agent-safehouse
Sandbox your local AI agents so they can read/write only what they need
This tool helps macOS developers prevent their local AI coding agents from accessing sensitive files or unintended parts of their computer. You tell it which specific folders and files your agent needs to do its work, and it ensures the agent can only read from or write to those approved locations. This is designed for developers who use AI tools for coding assistance and want to minimize risks.
1,184 stars. Actively maintained with 110 commits in the last 30 days.
Use this if you are a developer using AI coding agents on macOS and want to restrict their access to only the project files they need, preventing accidental data leaks or unauthorized modifications.
Not ideal if you are looking for a perfect security solution against a determined attacker, as this is a hardening layer for practical least privilege, not an impenetrable security boundary.
Stars
1,184
Forks
40
Language
Shell
License
Apache-2.0
Category
Last pushed
Mar 12, 2026
Commits (30d)
110
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/agents/eugene1g/agent-safehouse"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Featured in
Compare
Related agents
e2b-dev/E2B
Open-source, secure environment with real-world tools for enterprise-grade agents.
alibaba/OpenSandbox
OpenSandbox is a general-purpose sandbox platform for AI applications, offering multi-language...
e2b-dev/infra
Infrastructure that's powering E2B Cloud.
always-further/nono
Secure, kernel-enforced sandbox CLI and SDKs for AI agents. Capability-based isolation with...
boxlite-ai/boxlite
Sandboxes for every agent. Embeddable, stateful, snapshots, and hardware isolation.