skills.rest
etapsc
etapscCOMMUNITY
Explore all skills by this author →
GitHub

bridge-dual-agent

Coordinate Claude Code and Codex roles with markdown handoff files.

Updated Aug 23, 2026
Search Tags:#workflow#collaboration#codex#handoff#claude-code#dual-agent
One-click install
npx skills add https://github.com/etapsc/bridge --skill bridge-dual-agent
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: bridge-dual-agent
Source: https://github.com/etapsc/bridge/tree/main/bridge-dual-agent/.claude/skills/bridge-dual-agent
Command: npx skills add https://github.com/etapsc/bridge --skill bridge-dual-agent

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill streamlines collaboration between Claude Code and Codex by defining clear roles and communication protocols, ensuring efficient task handoffs and audits in a dual-agent development environment.

Core Features & Use Cases

  • Role Definition: Clearly assigns responsibilities between Claude Code (Orchestrator/Auditor) and Codex (Implementor).
  • Handoff Mechanism: Manages the creation of briefing documents (current-task.md) for Codex and the review of findings (codex-findings.md) by Claude Code.
  • Use Case: When working on a complex feature, use this Skill to ensure Claude Code plans the slice and audits Codex's implementation, preventing scope creep and ensuring quality.

Quick Start

Use the bridge-dual-agent skill to brief Codex before it begins implementing the current task.

Frequently Asked Questions about bridge-dual-agent

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

FAQPage Schema
How do I coordinate Claude Code and Codex for collaborative AI development?▼

You can coordinate Claude Code and Codex by defining distinct roles for orchestration, auditing, and implementation, using markdown files for structured task handoffs and findings within the BRIDGE framework.

What is the best way to manage task handoffs between Claude Code and Codex?▼

The best way to manage handoffs is by creating briefing documents like `current-task.md` for Codex to implement, and reviewing `codex-findings.md` to audit the results, preventing scope creep and ensuring quality.

When do I need a dual-agent protocol for AI development workflows?▼

You need a dual-agent protocol when working on complex features that require one AI to plan the task slice and audit implementation, while the other AI executes the code, ensuring efficient and structured collaboration.

Can I use Claude Code as an orchestrator while Codex acts as the implementor?▼

Yes, Claude Code can be assigned the Orchestrator and Auditor roles to plan slices and review findings, while Codex is assigned the Implementor role to execute the specified development tasks.

Does the bridge-dual-agent workflow require specific commands for briefing and gating?▼

Yes, the bridge-dual-agent workflow supports specific commands for briefing Codex before it begins implementing the current task and for gating progress within the structured BRIDGE framework.

Related Skills

Explore More Software Engineering Skills→
BeltDoorBeltDoor

codex-ready

Convert Claude Code repositories into Codex-compatible environments with adapter files and mirrored skills.

Community
Advanced
cskworkcskwork

review-codex

Automate final code or plan review with Codex CLI and write results to .task/review-codex.json.

Community
Advanced
ThiagoVitaleThiagoVitale

gsd-complete-milestone

Archive completed milestones and update project documentation for the next version.

Community
Advanced
ever-oliever-oli
3

opencode

Orchestrate autonomous coding tasks with the OpenCode CLI agent.

Community
Advanced