01.software Docs

Implementation Handoff

Choose SDK for application code, CLI for shell-capable agents, and narrower surfaces only when those defaults do not fit.

Implementation Handoff

Use this page after a workspace owner has approved the feature scope, credential owner, and launch path. Application runtimes should start with the SDK. Shell-capable coding agents should start with the Agent CLI. Use direct HTTP as an escape hatch and hosted MCP only for shell-less discovery.

Choose An Entry Point

Handoff Order

  1. Confirm enabled features in Feature Planning.
  2. Confirm key ownership, allowed origins, webhook ownership, and AI-agent scope in Integrations & Keys.
  3. Confirm launch criteria in Launch Readiness.
  4. Choose a project path in Build Options, then select the surface below at a routing level.
  5. Return to Launch Readiness before production exposure.

Surface Selection

NeedRoute toKeep out of this page
browser reads or product UISDK after Publishable Key and allowed origins are approvedSDK import tables or generated type details
privileged server workserver-side SDK, or API when the SDK does not fitendpoint-by-endpoint reference claims
shell-capable coding agentCLI after scope and credentials are approvedinteractive login or unstable text parsing in automation
hosted or shell-less AI clientMCP for three read-only discovery toolscollection CRUD or commerce workflows through MCP
event processingwebhooks after owner, retry, and observability responsibilities are assignedprovider-specific delivery guarantees not verified here

Do not implement against a workspace before the workspace owner approves scope, credential ownership, and launch readiness criteria.

Next Actions

  • New integration: open Authentication & Keys.
  • New or existing application project: choose Build Options.
  • App or product surface: start with SDK; use API when the SDK does not fit.
  • Shell-capable coding agent: open CLI.
  • Hosted, shell-less client: open MCP for read-only discovery.

On this page