Onboarding Configurations

Config Onboarding Skill

Automate the process of externalizing hardcoded paths, URLs, and environment-specific parameters into the project's configuration system.

Purpose

Ensure compliance with the Single Point of Truth (SPoT) architecture and Dynamic Shell Execution Protocol.

Process

Placeholder content for Process to satisfy validation.

1. Identify Hardcoded Parameters

Scan files for:

• Absolute paths (e.g., C:\..., /usr/bin/...)
• Hardcoded API or documentation URLs (e.g., https://github.com/...)
• Machine-specific identifiers

2. Externalize to tools.json

For each identified parameter:

1. Define a descriptive variable name (e.g., {NEW_VAR_PATH}).
2. Add the variable to .agent/config/tools.json with the hardcoded value as the default_url or in fallback_paths.
3. Map it to an environment variable in env_var.

3. Update .agentrules

Add the new variable to the ### Configuration Variables section in .agentrules using the {VARIABLE_NAME} = {VARIABLE_NAME} syntax.

4. Replace in Source

Replace the hardcoded value in the original file with the dynamic placeholder {VARIABLE_NAME}.

5. Verify Resolution

Ensure the agent can correctly resolve the new variable from the configuration hierarchy.

Rules

• No Hardcoding: Every absolute value must move to config.
• LLM Model Codes: Never hardcode LLM model names (e.g., gemini-2.0-flash). Always use ConfigManager to retrieve primary_model, preview_model, or fallback_model from settings.json.
• Defaults Matter: Always preserve the original hardcoded value as the fallback.
• Consistency: Use the same variable names across all files.
• Documentation: Update docs/CONFIGURATION.md if a new core variable is introduced.

When to Use

Placeholder content for When to Use.

Prerequisites

Placeholder content for Prerequisites.

Best Practices

Placeholder content for Best Practices.