mlflow/skills
9 skills2.9k installsGitHub
Install
npx skills add https://github.com/mlflow/skillsSkills in this repo
1Searching Mlflow Docssearching-mlflow-docs helps agents query MLflow official documentation for APIs, configuration, deployment options, and best practices while building ML infrastructure. It keeps implementation aligned with current MLflow guidance during tracking, registry, and serving setup without leaving the development workflow.414installs2Agent EvaluationMLflow agent-evaluation skill teaches Claude to create evaluation datasets, configure scorers, run batch agent benchmarks, and interpret MLflow metrics so LLM agents meet quality bars before production release.388installs3Instrumenting With Mlflow Tracinginstrumenting-with-mlflow-tracing teaches how to embed MLflow trace collection in Python ML and LLM workflows: defining spans around model inference and agent steps, attaching metadata, linking runs to experiments, and using traces to debug failures and compare pipeline behavior across versions.380installs4Mlflow Onboardingmlflow-onboarding walks new users through MLflow fundamentals—projects, experiments, tracking URIs, logging APIs, and registry basics—so teams can instrument ML code quickly. It targets the build stage when adopting standardized experiment management and reproducible model lifecycle practices.378installs5Analyzing Mlflow Traceanalyzing-mlflow-trace skill guides Claude through MLflow trace inspection—span hierarchies, tool failures, latency, and token costs—so operators debug agent regressions and improve orchestration in production agent systems.356installs6Querying Mlflow MetricsGuides agents to query MLflow tracking servers for run metadata, logged metrics, and experiment comparisons so teams can audit training history and monitor model quality over time.350installs7Analyzing Mlflow Sessionanalyzing-mlflow-session enables deep inspection of MLflow experiment sessions including runs, metrics, parameters, and artifacts to evaluate model quality and training stability. It supports ML teams in the operate stage when monitoring experiments and deciding whether models are ready for promotion or need further iteration.349installs8Mlflow Agentmlflow-agent from MLflow skills guides construction of traceable AI agents: configuring MLflow to log agent sessions, capturing tool invocations and outcomes, tying experiments to registry models, and establishing repeatable evaluation loops for agent behavior before ship and grow stages.241installs9Retrieving Mlflow Tracesretrieving-mlflow-traces is a Claude Code skill for ai & agent building. It helps solo builders move faster with AI-assisted coding.1installs