ralphex-settings
Scannednpx machina-cli add skill salmundani/ralphex/ralphex-settings --openclawRalphex Settings
Manage the ralphex configuration stored in .claude/ralphex.local.md.
Steps
-
Check if
.claude/ralphex.local.mdexists. If it does, read it and parse the YAML frontmatter to extract the current values ofbase_branch,codex_model, andcodex_reasoning_effort. If it does not exist, use these defaults:base_branch=main,codex_model=gpt-5.3-codex,codex_reasoning_effort=high. -
Show the user their current settings:
**Current ralphex settings:** - base_branch: {value} - codex_model: {value} - codex_reasoning_effort: {value} -
Ask the user which settings they want to change using AskUserQuestion.
-
For each selected setting, ask for the new value using AskUserQuestion:
- For
base_branch: suggest the current value plusmain,master,developas options. - For
codex_model: suggestgpt-5.3-codex,gpt-5.2-codex,gpt-5.3-codex-sparkas options. - For
codex_reasoning_effort: offerlow,medium,high,xhighas options withhighfirst.
- For
-
Write the updated settings to
.claude/ralphex.local.mdwith YAML frontmatter:--- base_branch: {value} codex_model: {value} codex_reasoning_effort: {value} --- -
Confirm the changes to the user by showing the new settings.
Source
git clone https://github.com/salmundani/ralphex/blob/main/ralphex/skills/ralphex-settings/SKILL.mdView on GitHub Overview
Ralphex settings lets you view and modify the core configuration stored in .claude/ralphex.local.md, including base_branch, codex_model, and codex_reasoning_effort. It helps keep runtime behavior aligned with your workflow by applying an interactive update flow, with defaults applied when the file is absent. Keeping these values in sync can improve model selection and branching behavior.
How This Skill Works
The skill reads YAML frontmatter from .claude/ralphex.local.md (or uses defaults if missing), displays the current values, and prompts you to change each setting with constrained options using AskUserQuestion. After you confirm, it writes the updated values back to the same file in YAML frontmatter and confirms the new settings.
When to Use It
- When you want to switch base_branch for testing in a different environment (e.g., main to develop).
- When you need to upgrade the codex_model to a newer variant like gpt-5.3-codex.
- When you want to adjust codex_reasoning_effort to balance speed and accuracy.
- When setting up a new environment and you want sensible defaults applied automatically.
- When auditing or documenting configuration changes for the team.
Quick Start
- Step 1: Run ralphex-settings to view current values.
- Step 2: Use AskUserQuestion prompts to select settings to change and provide new values from the allowed options.
- Step 3: Confirm updates; the skill writes the YAML frontmatter to .claude/ralphex.local.md and shows the new settings.
Best Practices
- Check if .claude/ralphex.local.md exists before editing to avoid overwriting defaults.
- Validate your input against the allowed option sets for each field.
- Back up the current file before applying updates.
- Document changes after updating so teammates understand the config.
- Update one setting at a time to minimize drift and confusion.
Example Use Cases
- Upgrade codex_model to gpt-5.3-codex after a release and test the behavior.
- Switch base_branch to develop for feature work and CI iterations.
- Lower codex_reasoning_effort to low for rapid prototyping.
- Reset to defaults by setting base_branch to main, codex_model to gpt-5.3-codex, codex_reasoning_effort to high.
- Log and share changes with the team to keep configs synchronized.