test-tui
openai/codex
Interactive testing guide for Codex TUI with debugging support.
What is test-tui?
Guide for starting and testing the Codex TUI interactively. Use this to verify changes and debug issues by running the TUI with trace logging and custom log directories.
- Start Codex TUI interactively for manual testing
- Enable trace-level logging with RUST_LOG environment variable
- Write logs to a custom directory for debugging
- Send test messages programmatically with proper sequencing
- Run via the just codex target with custom arguments
How to install test-tui
npx skills add https://github.com/openai/codex --skill test-tui- Codex repository cloned locally
- just task runner installed
- Rust environment configured
How to use test-tui
- 1.Set RUST_LOG=trace environment variable
- 2.Run just codex -c log_dir=<temp_directory> to start with logging
- 3.Interact with the TUI manually to test features
- 4.When sending test messages programmatically, send text first, then send Enter separately
- 5.Check the log directory for trace output to debug issues
Use cases
- Verify TUI changes before committing
- Debug TUI behavior with detailed trace logs
- Test message handling and input processing
- Troubleshoot interactive features in isolation
- Capture logs to a temporary directory for analysis
- Codex developers
- QA engineers testing TUI functionality
- Contributors debugging interactive features
test-tui FAQ
Sending text and Enter in one burst can cause issues with message processing. Send the text first, then send Enter in a separate write operation.
Pass the -c log_dir=<some_temp_dir> argument when running just codex to write logs to a specific directory.
Always set RUST_LOG=trace when starting the process to get detailed trace-level logging for debugging.
Use the just codex target, optionally with arguments like -c log_dir=<temp_dir>.
Full instructions (SKILL.md)
Source of truth, from openai/codex.
name: test-tui description: Guide for testing Codex TUI interactively
You can start and use Codex TUI to verify changes.
Important notes:
Start interactively.
Always set RUST_LOG="trace" when starting the process.
Pass -c log_dir=<some_temp_dir> argument to have logs written to a specific directory to help with debugging.
When sending a test message programmatically, send text first, then send Enter in a separate write (do not send text + Enter in one burst).
Use just codex target to run - just codex -c ...
Related skills
More from openai/codex and the wider catalog.
babysit-pr
Continuously monitor GitHub PRs, diagnose CI failures, handle review feedback, and auto-fix branch issues until merged or user help needed.
aspnet-core
Build, review, and architect ASP.NET Core web applications using current Microsoft guidance.
chatgpt-apps
Build ChatGPT Apps SDK applications with MCP servers and widget UIs using a docs-first workflow.
cli-creator
Build a durable, composable CLI from API docs, OpenAPI specs, or existing tools for Codex to run from any repo.
cloudflare-deploy
Deploy applications and infrastructure to Cloudflare using Workers, Pages, and platform services.
define-goal
Help the user define a concrete, measurable goal before starting work, especially when they ask to use the goal tool, create a goal, set an objective, clarify success criteria, or turn a fuzzy intention into a quantitative outcome. Use this skill for goal creation and goal refinement only; it does not manage durable snapshots, decision logs, or long-running execution artifacts.