菜单
Intent-scoped fabio skill for Fabric low-code data preparation: Dataflows Gen2 (Power Query / M mashup) and datamarts. Use for visual/low-code ETL that loads into a lakehouse or warehouse. Triggers: "dataflow", "dataflow gen2", "power query", "mashup", "low-code etl", "datamart", "transform without code".
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
Intent-scoped fabio skill for Fabric knowledge/graph and digital-twin modeling: ontology items (entity/relationship types and bindings), graph models, graph querysets, and Digital Twin Builder models/flows. Use to define/evolve ontologies, query graphs for agent grounding, and build operational digital twins. fabio can also export a tenant scan as OWL (context tenant --format owl) and import it. Triggers: "ontology", "fabric iq ontology", "knowledge graph", "graph model", "graph query", "entity type", "relationship type", "digital twin", "digital twin builder", "owl".
Intent-scoped fabio skill for Fabric CI/CD: stateless content-hash deploy (export/validate/plan/apply), Git integration, deployment pipelines, and variable libraries for environment-specific config. Use for promoting Fabric items between environments, Git lifecycle, and parameterized deployments. Triggers: "deploy", "ci/cd", "promote to production", "deploy plan", "deploy apply", "git commit", "git pull", "deployment pipeline", "variable library", "value set".
Intent-scoped fabio skill for the Fabric BI layer: semantic models (datasets), Power BI reports, paginated reports, and dashboards. Use for creating/refreshing semantic models, running DAX, and managing report items. Triggers: "semantic model", "dataset", "dax query", "refresh dataset", "power bi report", "paginated report", "dashboard", "direct lake".
Step-by-step release workflow for fabio. Invoke when cutting a new release version. Covers: version bump, dependency freshness, documentation updates, validation, changelog, tagging, and post-release dev version bump.
Intent-scoped fabio skill for Fabric administration and workspace management: workspace lifecycle/roles/folders/networking/encryption, capacity lifecycle, tenant-wide inventory and settings, domains, gateways, connections, managed private endpoints, and sensitivity labels. Use for governance, connectivity, capacity, and workspace-level administration. Triggers: "workspace", "assign capacity", "workspace roles", "network policy", "capacity", "resume capacity", "suspend capacity", "tenant settings", "domain", "gateway", "connection", "private endpoint", "sensitivity label".
Intent-scoped fabio skill for building application and API workloads on Microsoft Fabric: AI data agents (natural-language Q&A over data), GraphQL data APIs, User Data Functions (serverless), app backends (Power Apps), Cosmos DB databases (NoSQL OLTP), and Organizational Apps for distribution. Use to expose, query, and package Fabric data as apps/APIs/agents. Triggers: "data agent", "graphql api", "user data function", "serverless function", "app backend", "power app", "cosmos db", "organizational app", "publish app", "nl q&a".
| name | fabio-dataflows |
| description | Intent-scoped fabio skill for Fabric low-code data preparation: Dataflows Gen2 (Power Query / M mashup) and datamarts. Use for visual/low-code ETL that loads into a lakehouse or warehouse. Triggers: "dataflow", "dataflow gen2", "power query", "mashup", "low-code etl", "datamart", "transform without code". |
| license | MIT |
Generated file — do not edit by hand. This intent-scoped sub-skill of the
fabioskill is generated from fabio's command schema plus authored judgment. Regenerate withcargo test generate_subskills -- --ignored. For install, auth, output envelope, global flags, and agent-safety rules, see the rootfabioskill.
Prefer runtime introspection. This index is a snapshot; the installed binary is always authoritative. Use
fabio context agent --group <group>andfabio context describe <group> <command>for exact flags and output shapes.
Generated from fabio's command schema. For full flag details use fabio context agent --group <group> or fabio context describe <group> <command>.
Manage dataflows (Power BI data transformation)
| Command | Mutates | Description |
|---|---|---|
fabio dataflow create | yes | Create a new dataflow |
fabio dataflow delete | yes | Delete a dataflow |
fabio dataflow discover-parameters | no | Discover parameters of a dataflow |
fabio dataflow execute-query | no | Execute a query against a dataflow (returns Apache Arrow IPC) |
fabio dataflow get-definition | no | Get the definition of a dataflow |
fabio dataflow list | no | List dataflows in a workspace |
fabio dataflow run | yes | Run a dataflow on demand |
fabio dataflow show | no | Show details of a dataflow |
fabio dataflow update | yes | Update dataflow properties (name and/or description) |
fabio dataflow update-definition | yes | Update the definition of a dataflow |
Manage datamarts (Power BI)
| Command | Mutates | Description |
|---|---|---|
fabio datamart list | no | List datamarts in a workspace |
| Symptom | Fix |
|---|---|
| Unsure whether to use a dataflow or a pipeline | Transform data -> dataflow; sequence activities -> data-pipeline. Run 'fabio context disambiguate dataflow'. |
| Definition rejected | Ensure both the mashup and queryMetadata parts are present and valid. |
Cross-cutting operational guidance (the "common" layer) — consult the relevant topic before non-trivial work:
| Reference | Covers |
|---|---|
fabio context best-practices throttling | fabio transparently handles 429 (Too Many Requests) and gateway errors. Agents do NOT need to implement retry logic. |
fabio context best-practices pagination | fabio handles pagination via --all (auto-fetch all pages), --continuation-token (resume), and --limit (truncate). Agents rarely need to paginate manually. |
fabio context best-practices lro | Many Fabric operations are async (return 202). fabio polls them automatically. Use --wait for job operations. |
