{"id":"linuxserver-hydra","name":"hydra","af_score":44.2,"security_score":76.5,"reliability_score":35.0,"what_it_does":"Hydra is a configuration management framework (commonly for ML/AI and research codebases) that composes configurations, supports hierarchical config structures, and makes it easy to run experiments with parameter sweeps via command-line overrides.","best_when":"You need flexible, composable configuration and reproducible experiment runs, particularly in codebases that already follow a config-driven architecture.","avoid_when":"You require a networked configuration service with strong multi-tenant isolation, or you need built-in secret storage/rotation guarantees.","last_evaluated":"2026-03-30T13:43:50.186679+00:00","has_mcp":false,"has_api":false,"auth_methods":[],"has_free_tier":false,"known_gotchas":["Agents may incorrectly assume Hydra provides a network API (it primarily manages local configuration and launches runs).","Config composition/overrides can be subtle; agents should follow the repo’s configuration conventions and examples closely.","If using YAML/structured configs, agents should validate and type-check configs to avoid runtime failures."],"error_quality":0.0}