
Use case
Third-Party Script Shield
Run community and GitHub PowerShell with OS-level blast-shield isolation beside signing and approval.
- Pilot on one tenant
- Export audit slice
- Attach to customer SSP
Run community and GitHub PowerShell with OS-level blast-shield isolation beside signing and approval.
Visual summary before detailed use case copy.

Use case
Run community and GitHub PowerShell with OS-level blast-shield isolation beside signing and approval.
Problem: MSPs inherit scripts from forums, GitHub, and peer shares. Signing proves integrity at dispatch; sandboxing limits what the process can do at runtime.
Trustholm process sandboxing is opt-in (AllowProcessSandboxing). Assign platform:third-party-readonly-v1 to community scripts: no child processes, one active process, a memory cap, debug privilege stripped and LSASS access blocked.
Per-process network filtering is on the roadmap, not shipped. This is a same-kernel boundary, not VM isolation: it reduces blast radius.
Credentials still answer who ran the script; the sandbox limits what the process can do.
Pair with code signing and approval workflow. Export sandboxReport fields in security audit for assessor review.
No. Sandboxing is a blast shield beside Authenticode and tenant signing policy. Both layers are recommended for third-party content.