SenCLI · Developer and agent interface

One structured interface for people, pipelines, and agents.

SenCLI exposes the same platform and device semantics to an interactive terminal, automation, CI/CD, and MCP-capable agents—without giving automation a hidden side door.

sen · project workflow
$ sen project info greenhouse-edge
project      greenhouse-edge
status       linked

$ sen project app validate
✓ application source is valid

$ sen project build trigger --wait
QUEUED → RUNNING → SUCCEEDED
Structured output is available for automationexit 0

Same contract, different operators

Automation should inherit the guardrails—not bypass them.

SenCLI keeps validation, identity, permissions, and failure semantics aligned across every operator.

01

Developer terminal

Inspect a project, validate definitions, build firmware, and operate devices with discoverable commands and explicit feedback.

02

CI/CD pipeline

Use deterministic exit status and structured output while preserving the same policy and release identities.

03

AI agent via MCP

Give an agent constrained tools with typed inputs and observable results instead of unrestricted shell or private API access.

Predictable automation

The same checks, whether a person or an agent runs the command.

SenCLI keeps project context, permissions, validation, and results consistent. Teams can automate routine device work without creating a separate, harder-to-audit path.

REQUESTCommand with explicit parameters
↓ verify
GUARDRAILSProject context · permission · validation
↓ run
RESULTReadable feedback + automation-ready output

Build an automation surface you can reason about.

Use the same capability and operation model in a terminal, pipeline, or agent workflow.