Reasoning layer
Use MCP and docs to plan: discover what needs to happen, identify the right tenant, and decide the remediation or admin action.
The CLI is designed to be predictable for humans, shell scripts, CI jobs, and coding agents: stable resource-action verbs, clear flags, JSON output, and tenant-aware execution.
Sign in, choose the active tenant, inspect devices, and trigger a remediation run with a human-reviewed command sequence.
smplify auth login
smplify tenant list
smplify tenant use <tenant-id>
smplify device search --os windows --status vulnerable --output json
smplify remediation run patch-critical --device dev-123
Prefer JSON output in scripts so downstream tooling can filter, transform, and act on results without rebuilding API auth and request handling.
smplify device search --os windows --status vulnerable --output json \
| jq '.[] | select(.risk_status=="vulnerable") | .id'
For tools like Codex or Claude Code, the CLI becomes the deterministic execution layer underneath a higher-level reasoning loop.
Use MCP and docs to plan: discover what needs to happen, identify the right tenant, and decide the remediation or admin action.
Use the CLI to execute: the command structure is stable, scriptable, and easier for agents to call repeatedly and safely.
The CLI is optimized for composability and predictable execution, not for interactive hidden state.
smplify <resource> <action> is the core command pattern.
Machine-readable output is available wherever automation needs it.
Active tenant context is persisted and injected into requests consistently.
Installers resolve published binaries and checksums from GitHub Releases.