An advanced project management layer for OpenAI Codex CLI that automates task execution, monitoring, and quality assurance.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install codex-agent-enhanced
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-agent-enhanced 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 Agent Enhanced?
Codex Agent Enhanced is a comprehensive management system designed to elevate the OpenAI Codex CLI from a simple command-line tool to a proactive project manager. It enables developers to handle complex workflows involving knowledge base maintenance, task orchestration, and real-time monitoring through tmux-based TUI sessions. By integrating seamlessly with Openclaw Skills, this agent handles the heavy lifting of requirement analysis, prompt design, and iterative quality checks.
This skill is particularly effective for developers who need to manage multi-step coding tasks with high autonomy. It utilizes notify hooks and pane monitors to provide asynchronous updates, ensuring that users are only alerted for critical decisions or final reports. Whether you are performing routine refactoring or architecting new features, this skill ensures that every Codex interaction is tracked, validated, and aligned with project standards.
Codex Agent Enhanced Use Cases
- Automating multi-step software development tasks using the OpenAI Codex CLI.
- Managing long-running coding sessions with asynchronous notifications via Telegram.
- Maintaining a synchronized knowledge base of AI features, model schemas, and prompting patterns.
- Implementing project-level isolation for AI agents using localized environment configurations.
- Performing automated code reviews and quality gate checks before committing changes.
How Codex Agent Enhanced Works
- Requirement Analysis: The agent identifies task goals, acceptance criteria, and relevant technical stacks by consulting localized knowledge files.
- Strategy Selection: Based on task complexity, the agent chooses between execution modes (single-command exec vs. multi-turn TUI) and selects the optimal AI model.
- Environment Isolation: Project-specific environment variables are loaded from a .env file to ensure concurrent tasks do not conflict.
- Task Execution: The agent launches the Codex CLI, either directly or within a managed tmux session, and monitors the output.
- Automated Monitoring: Background hooks (pane_monitor.sh or on_complete.py) track progress and trigger wake-up signals for the agent when intervention or review is needed.
- Quality Assurance: The agent evaluates the results against a state-driven quality standard, performing autonomous fixes for style or minor logic issues.
- Reporting: A final summary is generated, detailing changes, intermediate iterations, and any architectural deviations for user approval.
Codex Agent Enhanced Setup
To integrate this skill into your workflow, follow these installation steps:
- Install tmux on your local system if not already present.
- Update your
~/.codex/config.tomlto include the notify hook:
notify = ["python3", "/path/to/skill/hooks/on_complete.py"]
- Configure the required environment variables in your project root
.envfile:
OPENCLAW_AGENT_NAME="your-agent-name"
OPENCLAW_AGENT_CHAT_ID="your-telegram-id"
- Verify that the Codex CLI is accessible and that the Openclaw Skills messaging system is functional by running
codex --version.
Codex Agent Enhanced Data Schema & Taxonomy
The skill maintains a structured data hierarchy to track task state and knowledge:
| Component | File Path | Purpose |
|---|---|---|
| Task State | .codex-task-state.json |
Stores active task IDs, status (running, pending, blocked), and notification metadata. |
| Knowledge Base | knowledge/features.md |
Tracks CLI capabilities and available feature flags. |
| Configuration | knowledge/config_schema.md |
Defines the validation rules for the Codex TOML configuration. |
| Versioning | state/version.txt |
Monitors the current Codex version for automated knowledge updates. |
Codex Agent Enhanced Advanced Features
- Asynchronous Wakeup: Uses notify hooks to alert the agent only when Codex completes a turn or requires manual approval.
- Pane Monitoring: Real-time tmux output scanning to detect approval prompts in non-automated modes.
- Project Isolation: Dedicated .env support allows multiple AI-driven projects to run concurrently without parameter collisions.
- Automated Knowledge Updates: Protocol-driven workflows to track and incorporate new AI model features and CLI changes.
- Multi-Agent Support: Integration with Openclaw Skills cron jobs for persistent heartbeat monitoring and task resumption.
SKILL.md
Loading
Related Openclaw Skills
Ahrefs SEO Analysis
A professional-grade SEO integration for analyzing backlinks, keyword rankings, and technical site health via the Ahrefs v3 API.
NodeTool
A local-first visual AI workflow builder that combines ComfyUI node flexibility with n8n automation power for LLM agents and RAG pipelines.
Polymarket Optimizer
An automated parameter tuning engine that analyzes trading performance and adjusts strategy thresholds to maximize prediction market returns.
Polymarket Executor
An autonomous multi-strategy trading bot designed to scan and execute trades across all Polymarket prediction categories.
Cron Doctor
An automated diagnostic agent for identifying, triaging, and fixing cron job failures across Linux and macOS environments.
Agent Task Logger
A real-time task logging system for AI agents that automatically detects workspaces and tracks execution progress, commands, and status.
Featured*
