A methodical end-to-end software delivery orchestrator for Codex CLI that manages project lifecycles through strict quality gates and spec-driven development.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install codex-conductor
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install codex-conductor using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Codex Orchestrator?
Codex Orchestrator transforms AI coding agents from simple generators into disciplined delivery systems. It operates on the non-negotiable principle of spec-driven development, ensuring that no implementation occurs without a pre-approved specification. By integrating this skill into your workflow, you leverage one of the most structured Openclaw Skills available for professional software engineering.
The system supports both greenfield projects for building new systems and brownfield projects for modernizing existing codebases. It manages the entire lifecycle—from initial planning and architecture to vertical slice implementation and security gates—ensuring high-quality output through rigorous verification loops and automated progress tracking.
Codex Orchestrator Use Cases
- Launching new software projects with a comprehensive greenfield delivery framework.
- Onboarding and modernizing legacy systems using the brownfield modernization workflow.
- Implementing complex features through vertical slices that require mandatory verification against specs.
- Managing multi-agent development environments where human-in-the-loop approvals are necessary at critical gates.
- Ensuring strict documentation and traceability for enterprise-grade AI-assisted coding projects.
How Codex Orchestrator Works
- The orchestrator begins with an intake process, utilizing a planning questionnaire to define mission, scope, and technical requirements.
- A detailed specification is authored and approved, serving as the mandatory foundation for all subsequent code implementation.
- The system scaffolds the project environment, creating the AGENTS.md contract and initializing the .orchestrator status engine.
- Coding agents execute tasks in single-task increments, directly updating documentation and progress logs upon completion.
- The Openclaw Skills verification engine runs automated CLI checks or browser-based UI reviews to validate output against the spec.
- Project progression is managed through a sequence of gates (G0-G7), where each transition requires validated evidence and status updates.
Codex Orchestrator Setup
To initialize a project and prepare the necessary documentation artifacts, run the following command in your terminal:
python scripts/init_project_docs.py --root <project-path> --mode <greenfield|brownfield>
Ensure you have configured the selected coding agent and defined your execution mode (autonomous or gated) within the generated context files before triggering the first gate.
Codex Orchestrator Data Schema & Taxonomy
Codex Orchestrator maintains a rigorous data schema to ensure project traceability and machine-readable state management:
| Artifact | Format | Purpose |
|---|---|---|
AGENTS.md |
Markdown | The project workflow contract for all participating coding agents. |
.orchestrator/status.json |
JSON | Machine-readable state tracking gate progress and compliance. |
docs/progress.md |
Markdown | Human-readable log of completion percentages and current milestones. |
docs/traceability.md |
Markdown | Mapping of code implementations back to their original specifications. |
docs/change-log.md |
Markdown | Continuous record of all modifications and their associated impact analysis. |
Codex Orchestrator Advanced Features
- Dual execution modes offering a choice between fully autonomous delivery or gated human-approval cycles.
- Change impact management via a dedicated script that identifies documentation and code areas affected by new requirements.
- Support for multi-agent workflows, allowing seamless switching between models like Codex, Claude, or custom LLMs.
- Integrated verification loops where Openclaw Skills automatically retry failed tasks with specific error feedback for the agent.
- Comprehensive gate engine (G0-G7) that prevents progression until all security, quality, and documentation checklists are satisfied.
SKILL.md
Loading
Related Openclaw Skills
ClawCost
A financial tracking utility for AI agents that monitors token costs, budget limits, and model-specific expenditures.
FreeRide
FreeRide automates the selection and configuration of free AI models from OpenRouter for the OpenClaw agent ecosystem.
FreeRide
FreeRide is a management tool that automatically configures and ranks free AI models from OpenRouter for use within OpenClaw environments.
NEAR Name Service
A comprehensive utility for managing .near domain names and resolving blockchain account IDs within the Openclaw Skills ecosystem.
Async Web Search
A high-performance parallel web search tool using Python's asyncio to retrieve data across multiple queries simultaneously.
Agent Telegram Communication
A standardized communication protocol for AI agent teams to send structured status updates and notifications via Telegram.
Featured*
