executing-plans
executing-plans
Use when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available
configure-railsCreate and validate NVIDIA NeMo Guardrails configs: config.yml, Colang flows, prompts, guardrail catalog rails, custom actions, providers, caching, knowledge base, and config migration.
Independent GitHub snapshot · Automated classification · Original documentation governs
Use this sub-skill when the task is about authoring or validating a guardrails configuration, not running chat, server, evaluation, or source-checkout work.
config.yml, prompts.yml, rails.co/main.co, config.py, actions.py, kb/, import_paths, caching, or knowledge-base settings.RailsConfig.from_path() or RailsConfig.from_content().models/rails entries.nemoguardrails convert./v1/* calls: ../run-rails/SKILL.md.../evaluate-and-observe/SKILL.md.../repo-development/SKILL.md.../setup-and-basics/SKILL.md.references/configuration-and-colang.md to choose the right folder layout, loader, and Colang version.references/guardrail-catalog.md to map the desired rail family to the required flow names, prompt tasks, model types, and optional dependencies.references/custom-actions-and-providers.md for actions.py, config.py, provider registration, and embedding provider registration.references/troubleshooting.md whenever config loading fails or a validation rule is expected to fire.scripts/validate_config.py; pass --instantiate only when you want to confirm that LLMRails can be constructed without generating a response.python sub-skills/configure-rails/scripts/validate_config.py --config path/to/config
python sub-skills/configure-rails/scripts/validate_config.py --config path/to/config --instantiate
The helper loads the config, prints a concise summary, and never calls generation, chat, or server methods.