domino-modeling-assistant
Enable AI-assisted model development (vibe modeling) in Domino. The assistant writes the data-analysis, model and training code and uses the bundled Domino MCP server to run it as Domino Jobs, wait for and read their output, and read, upload and sync files (text or binary) for DFS projects, so work stays reproducible, governed and tracked in Experiment Manager. Use when setting up an AI coding assistant to work with Domino, configuring or troubleshooting the Domino MCP server, or when the user asks to analyze a dataset, train or compare models, or run scripts as Domino Jobs from a coding assistant.
- License 许可
- MIT
- 最近更新
- 2026-11-14
- Stars
- ★ 5
- Forks
- 4
01 前言 Front Matter
- name
- domino-modeling-assistant
- description
- Enable AI-assisted model development (vibe modeling) in Domino. The assistant writes the data-analysis, model and training code and uses the bundled Domino MCP server to run it as Domino Jobs, wait for and read their output, and read, upload and sync files (text or binary) for DFS projects, so work stays reproducible, governed and tracked in Experiment Manager. Use when setting up an AI coding assistant to work with Domino, configuring or troubleshooting the Domino MCP server, or when the user asks to analyze a dataset, train or compare models, or run scripts as Domino Jobs from a coding assistant.
- compatibility
- Domino 6.3 and Domino Cloud. Inside a Domino Workspace the bundled MCP server calls the platform through DOMINO_API_PROXY with no credentials; outside Domino it needs DOMINO_HOST and DOMINO_TOKEN (a Personal Access Token or service account token).
02 资源文件 2
03 SKILL.md 正文
Modeling assistant
Applies to Domino 6.3 and Domino Cloud.
Key components
MCP (Model Context Protocol)
The Domino MCP server bridges the coding assistant and the Domino platform APIs needed for running Jobs (analysis and training scripts run better as Jobs than on the local terminal), waiting for and reading their output, and reading, uploading and syncing files of a Domino File System (DFS) project when it does not use a Git repository.
The server is bundled with this plugin (mcp-servers/domino_mcp_server/, name domino_server) and starts automatically when the plugin is installed from its source repository. Skills-only distributions do not include it.
If the domino_server tools (such as run_domino_job) are not available in this session, do not pretend to call them. Tell the user, point them at SETUP.md, and meanwhile use the domino-jobs and domino-python-sdk skills to run work through the Domino REST API.
- Inside a Domino Workspace: nothing to configure. The server calls the platform through
DOMINO_API_PROXY(noAuthorizationheader; the proxy adds the user's token) and reads the project fromDOMINO_PROJECT_ID,DOMINO_PROJECT_OWNER,DOMINO_PROJECT_NAMEandDOMINO_IS_GIT_BASED. On hosts that start MCP servers with a stripped environment (Codex), it recovers these variables from its parent process. - Outside Domino (laptop, CI): set
DOMINO_HOST(the deployment's HTTPS URL) andDOMINO_TOKEN(a Personal Access Token or service account token, sent asAuthorization: Bearer); Domino user and Project environment variables are not readable there, so the host's configuration or the env file holds the laptop's copy. Hosts with plugin settings ask for both at install; the others read the env file~/.config/domino-mcp-server/.env(Codex passes the server no other location). Per-host steps: SETUP.md. Legacy user API keys (DOMINO_API_KEY,X-Domino-Api-Key) are deprecated and not accepted. See SETUP.md.
Tool failures come back as MCP tool errors whose message says what to fix; relay it to the user instead of retrying blindly.
Docs: https://docs.domino.ai/cloud/platform-capabilities/features/coding-assistants/modeling-assistant (6.3: replace cloud with 6.3).
Important considerations
Whenever possible run commands as Domino Jobs through the domino_server MCP tools rather than on the local terminal.
At the start of every session, call get_domino_environment_info. It tells you whether you are inside a Domino Workspace or on a laptop, the project owner, name and id, whether the project is Git-based or DFS, which authentication mode is active, and what is missing. Inside a Workspace everything is auto-detected; you do not need domino_project_settings.md. Outside Domino, fall back to domino_project_settings.md for the project name, user name, and dfs setting.
Waiting for a Job happens on the server, never in a loop and never with shell sleep (hosts block it): start with run_domino_job(..., wait_seconds=300), call check_domino_job_run_status(run_id, wait_seconds=300) again while the response says is_completed is false, then read the full output with check_domino_job_run_results (log_type="stdErr" when the status is not Succeeded), briefly explain the conclusions, and share any MLflow or experiment run URL (open_web_browser outside Domino, a link inside a Workspace). Caps and response fields: TOOLS.md.
Pass hardware_tier and environment_id to run_domino_job only when the user asks for a specific tier or environment; both are validated before the Job starts and an unknown value is an error that lists what exists. Omit both for the project defaults.
Assume project data is already in the Domino project and accessible from job runs, under /mnt/data/ or /mnt/imported/data/ (Git-based) or /domino/datasets/ (DFS); confirm the full path by running a job that lists the folder contents recursively. Always create scripts to understand and transform data and run them as jobs; print analytical output as plain-text tables on stdout so job results are easy to check. Do not delete analysis scripts; save summary charts as image files in the project.
Always check whether the local project has uncommitted changes. For Git-based projects, commit and push before running any Domino job, otherwise Domino cannot see the new files; the file tools refuse Git-based projects. For DFS-based projects, sync with the MCP file tools instead of git before running any jobs, preferring smart_sync_file, and never force an overwrite without telling the user what will be replaced. Conflict and overwrite rules, text versus binary content: TOOLS.md.
When training a model, use MLflow instrumentation and assume the tracking server is configured by Domino; do not set a tracking URI.
MCP server tools
get_domino_environment_info, run_domino_job, check_domino_job_run_status, check_domino_job_run_results, open_web_browser, list_domino_project_files, download_file_from_domino_project, upload_file_to_domino_project, sync_local_file_to_domino, smart_sync_file. Signatures, the waiting rules and caps, hardware tier and environment validation, and text versus binary file handling: TOOLS.md. Routes behind each tool: mcp-servers/domino_mcp_server/README.md in the plugin repository.
04 同仓库更多 Skills
tags-and-properties
Manage Domino tags and properties via the Tags (Taxonomy) API at /api/taxonomy/v1. Covers namespaces and tags (create, list, update, deprecate, delete; tag entities such as projects, models, datasets, apps, project templates, NetApp volumes, environments; query by tag; autocomplete; merge; CSV import/export) and properties (typed metadata fields - text, number, date, boolean, url, user, organization, select, multi_select - with per-entity values, groups and versioned entities like model_version/app_version). Use when organizing entities with tags, building hierarchical namespaces, defining typed metadata fields, setting property values, finding entities by tag, bulk-tagging during onboarding, or migrating a taxonomy across environments.
netapp-volumes
Work with Domino Volumes for NetApp ONTAP (NetApp Volumes) - shared, multi-terabyte storage with near-instant snapshots. Covers creating volumes, attaching them to projects, mount paths (/mnt/netapp-volumes/ or /domino/netapp-volumes/), snapshots and tags, roles, deletion, mounting volumes in jobs, and the NetApp Volumes REST API (/remotefs/v1). Use when managing large-scale shared data, taking fast no-copy snapshots, or integrating NetApp ONTAP storage with Domino.
domino-workspaces
Work with Domino Workspaces - interactive sessions running Jupyter, JupyterLab, VS Code or RStudio on a hardware tier with a compute environment. Covers launching and configuring a Workspace, what persists across stop and resume, file paths and syncing for DFS and Git-based Projects, package and home-directory persistence, SSH access from a local IDE, the Workspace API routes, and troubleshooting. Use when setting up or debugging an interactive development session in Domino.
domino-ui-bootstrap
Bootstrap or retrofit a Vite + React 18 + TypeScript project so it uses the Domino design system (`@dominodatalab/extensions-tools`). Scoped to a **fully standalone SPA frontend for a Domino application or extension**, not snippets, library additions, or work inside an existing Domino monorepo package. Use whenever the user wants to create, scaffold, refactor, retrofit, or set up such an SPA (a React app, web app, frontend, UI, or extension) that should "look like Domino", use Domino components, follow the Domino design system, or integrate with the Domino platform, including when they mention DominoThemeProviderDecorator, extensions-tools, or base-components without saying "Domino". Also use to make an existing standalone React/Vite project "Domino-styled" or add Domino theming. Handles React 18 / react-router 5 pinning, Storybook MCP registration, theme provider wiring, and build verification.
domino-trace-setup
Set up GenAI tracing for an LLM or agent application in Domino by checking the Domino SDK and MLflow requirements (the DSE needs nothing installed), then adding a tracing_setup.py helper, span evaluators, a config.yaml, and @add_tracing / DominoAgentContext wiring. Use when the user asks to add tracing, observability, or evaluation to an agent or LLM app running on Domino.
domino-rust-apps
Build a Rust web app (axum) in Domino and ship it as a Domino App that starts from a prebuilt binary. Covers installing a pinned Rust toolchain into a Compute Environment with rustup, Cargo project defaults (edition 2024, rust-toolchain.toml, Cargo.lock, Clippy pedantic, rustfmt, cargo-nextest, cargo-deny), rust-analyzer for the IDE and coding agents, check hooks, calling Domino as the viewer with extended identity propagation, dev and release build profiles, keeping target/ out of Git and Domino storage, and an app.sh that runs the binary in App Preview and deployments without compiling. Use when the user wants a Rust app, API, dashboard or Proa site on Domino, a performance-sensitive or correctness-critical App, or asks to set up Rust, cargo, Clippy, nextest or rust-analyzer in a Domino Workspace or environment.