serving-and-projects
Use when working with MLflow CLI commands, tracking servers, auth/security, database migrations, MLproject execution, local model serving/prediction, deployments, MCP or agent setup, and local development server workflows.
- License 许可
- Apache-2.0
- 最近更新
- 2026-10-08
- Stars
- ★ 11
- Forks
- 3
01 前言 Front Matter
- name
- serving-and-projects
- license
- Apache 2.0
- metadata
- { "disco-role": "operating" }
- description
- Use when working with MLflow CLI commands, tracking servers, auth/security, database migrations, MLproject execution, local model serving/prediction, deployments, MCP or agent setup, and local development server workflows.
- disable-model-invocation
- true
02 资源文件 4
- 参考文档references/cli-reference.md7.4 KB
- 参考文档references/projects-and-deployments.md7.5 KB
- 参考文档references/server-and-auth.md6.3 KB
- 参考文档references/troubleshooting.md6.9 KB
03 SKILL.md 正文
Serving And Projects
Use this sub-skill when the task is about operating MLflow from the command line or hosting/serving MLflow services. It covers command discovery, mlflow server, mlflow run, mlflow deployments, mlflow models serve/predict, auth app setup, database migrations, MCP/agent CLI setup, and development-server caveats.
Route First
- Use
references/cli-reference.mdto choose CLI groups, inspect help safely, and distinguish read-only probes from side-effectful commands. - Use
references/server-and-auth.mdfor tracking server storage, artifact serving, allowed-host/CORS security, auth app setup, and auth DB migration guidance. - Use
references/projects-and-deployments.mdforMLprojectexecution, deployment target commands, local model serving/prediction, Docker/SageMaker/cloud cautions, MCP, agent setup, and repo development server workflows. - Use
references/troubleshooting.mdwhen diagnosing backend/artifact store mismatches, auth DB problems, host/origin blocks, deployment credentials, env-manager failures, CLI input-file formats, or long-running process behavior.
Safety Defaults
- Prefer help/version probes before running operational commands:
python scripts/cli_probe.py --commands mlflow "mlflow server" "mlflow models serve"; use--python-module /path/to/pythonwhen themlflowconsole script is not onPATH. - Do not start
mlflow server,mlflow models serve,mlflow mcp run,mlflow gateway start, or the repo dev server unless the user explicitly asks for a long-running process and provides ports/log handling. - Do not run deployment create/update/delete, SageMaker, Docker build/run, Kubernetes, Databricks, or cloud commands unless credentials, target, network, and cost/side effects are explicitly approved.
- Do not run auth DB migrations or tracking DB migrations against shared databases without a backup and the exact database URI from the user.
- For model flavor internals, serialization, signatures, and environment inference, route to
models-and-flavors; for experiment/run semantics and registry objects, route totracking-and-registry; for GenAI tracing/evaluation semantics, route togenai-observability.
Safe Probe Script
scripts/cli_probe.py runs only --help and --version style probes with timeouts. It is suitable for checking an installed MLflow CLI surface before choosing commands, and it intentionally refuses arbitrary command execution.
04 同仓库更多 Skills
obsidian-clipper-template-creator
Guide for creating templates for the Obsidian Web Clipper. Use when you want to create a new clipping template, understand available variables, or format clipped content.
obsidian-clipper-template-creator
Guide for creating templates for the Obsidian Web Clipper. Use when you want to create a new clipping template, understand available variables, or format clipped content.
python-api
Programmatic Object-Detection-Metrics API guidance for BoundingBox, BoundingBoxes, Evaluator, enums, coordinate conversion, AP/mAP semantics, and source-style import diagnostics.
file-evaluation
Folder-based PASCAL VOC AP/mAP evaluation from ground-truth and detection text files.
object-detection-metrics
Guides legacy Object-Detection-Metrics PASCAL VOC AP/mAP evaluation from detection text folders and direct Python API classes.
nx-workspace-patterns
Configure and optimize Nx monorepo workspaces. Use when setting up Nx, configuring project boundaries, optimizing build caching, or implementing affected commands.