Skip to content
Adventive Adventive Engineering

Decision: Shell CLI — Nice-to-Have, Defer Post-Launch

Decision: Shell CLI — Nice-to-Have, Defer Post-Launch

Section titled “Decision: Shell CLI — Nice-to-Have, Defer Post-Launch”

Date: 2026-04-29
Status: Deferred
Author: Jeffrey Lambert / Claude

Context

Section titled “Context”

During Phase 6 implementation review, the idea of a shell CLI (command-line client for the Adventive Public API) was raised as a quality-of-life tool for developers and ops. A CLI would allow ad-hoc endpoint queries without maintaining Postman collections or writing curl one-liners.

Decision

Section titled “Decision”

Do not implement a CLI in this repo or in this migration phase.

Rationale

Section titled “Rationale”
  1. Out of scope. The migration plan (Phases 4–7) targets a Worker codebase, CI pipeline, smoke scripts, and runbook. A CLI is a separate development tool, not part of the API itself.
  2. Smoke scripts cover the ops need. scripts/smoke.sh already exercises all endpoints with structured assertions. A CLI would add little value for deployment verification.
  3. Postman collection covers the dev need. A Postman collection was created alongside the API as the primary exploration/testing tool. It already has pre-request auth scripting and saved environments for dev/stg/prd.

When to revisit

Section titled “When to revisit”

After the API is live in production and adoption grows, a lightweight CLI (TypeScript, distributed via npm or as a Cloudflare Worker-side tool) could improve DX for client-side integrators. Triggers to revisit: (a) multiple external integration partners, (b) repeated Slack requests for “how do I query X,” or (c) a dedicated platform SDK project is started.

A CLI would logically live in a separate adventive-api-cli repo and consume the public API as an external client — it has no reason to share code with this Worker repo.