What sandboxes provide
- Complete isolation: Each task runs in its own environment. Nothing can reach your host systems or network unless you explicitly configure it.
- Pre-loaded security tooling: Sandboxes come pre-configured with security tools, CLI utilities, compilers, and dependencies so Neo can start testing immediately.
- Controlled execution: Network access, system calls, and resource usage are monitored and controlled throughout every run.
- Real-time visibility: All execution logs, outputs, and artifacts stream back to your conversation as they happen, so you see exactly what Neo is doing.
How execution works
When Neo needs to run something, here’s what happens:- Environment spins up: A fresh sandbox initializes with the necessary tools and dependencies
- Credentials injected securely: Environment variables and secrets you’ve defined are injected at runtime, never written to disk
- Task executes: Neo runs commands, scripts, or tools within the isolated environment
- Results stream live: Output, logs, and progress stream back to you in real time
- Artifacts captured: All generated files, reports, screenshots, and evidence are automatically saved
- Environment destroyed: The ephemeral sandbox is destroyed after execution completes
Persistent storage for your artifacts
While the sandbox itself is ephemeral (destroyed after each run), all generated files and artifacts are persisted to a dedicated storage layer. This means:- Nothing gets lost: Every file created during execution (scan results, reports, screenshots, logs, compiled binaries) is automatically saved
- Build on previous work: Files generated in one task remain available for future tasks, so Neo can reference prior results and build on them
- Evidence preserved: Complete audit trails and artifacts from security assessments are retained for review, reporting, and compliance
- Workspace accumulates context: Your workspace maintains state across executions, so results compound over time

