| name | ngat-config |
| description | Manage and validate runtime configuration schemas and YAML files under config/ plus root config.yaml for the trading engine. Use when adding config fields, fixing env interpolation, or resolving mismatches between configured and supported timeframes/indicators. |
Ngat Config
Overview
Control how runtime settings are modeled and loaded. Use this skill when schema evolution or config correctness impacts startup and worker behavior.
Submodules
config/settings.py: Pydantic models for feed/broker/storage/strategy/settings.config/config.yaml: runtime config (symbols, base/derived timeframes, indicators).config/sample_config.yaml and root config.yaml: alternates/examples.
Pros
- Keep typed config contract through Pydantic.
- Keep feed/broker/storage abstraction explicit at configuration level.
Cons
- Keep timeframe validator too narrow for current YAML (
1d/1w/1M/1y unsupported by validator).
- Keep duplicate config files with drift risk.
Caveats
- Treat env interpolation as global and invasive in CLI loader.
- Keep naming consistent between
day vs 1d; mixed vocabulary causes runtime failures.
- Assume indicators list may be consumed by multiple services; avoid shape changes without fallback.
Forward Features
- Add schema versioning and migration helpers.
- Add strict timeframe enum shared by config, aggregator, API, and frontend.
- Add startup config lint command with actionable error messages.
