codex-sdk

Create durable, auditable Codex coding workflows with ExecPlans and SQLite-backed state.

5|Updated Jan 26, 2026
One-click install
npx skills add https://github.com/BjornMelin/dev-skills --skill codex-sdk
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: codex-sdk
Source: https://github.com/BjornMelin/dev-skills/tree/main/skills/codex-sdk
Command: npx skills add https://github.com/BjornMelin/dev-skills --skill codex-sdk

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

Build reliable, auditable, multi-step coding workflows using Codex that scale from a single run to multi-agent orchestration.

Core Features & Use Cases

  • Durable planning with ExecPlans to survive context folds and restarts across teams.
  • SQLite-backed memory and audit trails for reproducibility and governance.
  • MCP server orchestration with the OpenAI Agents SDK for multi-agent collaboration and gated handoffs.

Quick Start

Run Codex CLI or Codex SDK workflows to start a durable, auditable coding pipeline with an MCP server and Agents.

Frequently Asked Questions about codex-sdk

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build durable coding workflows that survive context folds and restarts?

Durable coding workflows use ExecPlans to maintain execution planning state, ensuring multi-step coding tasks resume correctly after interruptions or context window limits are reached.

What is MCP server orchestration for multi-agent coding collaboration?

MCP server orchestration coordinates multiple coding agents using the OpenAI Agents SDK, enabling gated handoffs where planners, developers, and testers collaborate with structured execution traces and safe defaults.

How do I create auditable coding pipelines with reproducible governance trails?

Auditable coding pipelines use SQLite-backed memory to store execution state and audit trails, providing reproducibility and governance by recording every step from planning through multi-agent handoffs.

Can I use Codex SDK workflows for both single-run tasks and multi-agent orchestration?

Codex SDK workflows scale from single-run coding tasks to multi-agent orchestration, providing structured outputs, MCP server coordination, and SQLite-backed state management across both execution patterns.

Does the Codex SDK approach work without external dependencies for state management?

The Codex SDK approach operates with no external dependencies, using SQLite-backed state management internally to provide memory, audit trails, and resumable workflow execution without requiring additional packages.

When should I use gated handoffs in multi-agent coding workflows?

Gated handoffs are used in multi-agent coding workflows when planners, developers, and testers need structured collaboration points with execution traces, ensuring safe defaults and reproducible state transitions between agents.