Menu
Injected at session start. Provides overview of the Alation plugin, skill routing, and guidance on working with users.
Instalar com Codex ou Claude Copie este prompt, cole no Codex, Claude ou outro assistente e deixe que ele revise a página da skill e instale para você.
Baseado na classificação ocupacional SOC
Use when the user wants to discover, find, or browse data assets in the Alation catalog, data products, or marketplaces. Triggers: search catalog, find tables, list data sources, browse schemas, show columns, what data do we have, what tables exist, explore database, describe table, what's in this data source, show me the schema, list schemas, list columns, what databases do we have, show table structure, describe this data source, find data products, list data products, what data products do we have, search products, show available products, what products are in the marketplace, browse marketplace, which product has sales data. Use this skill when the user wants to know what data exists, understand table/column structure, or find data products — not when they want actual query results or numbers. Not for answering data questions or running queries — use ask instead. Not for creating or managing data products — use curate instead.
Use when the user asks a data question, wants to run a query against a data product, chat with an AI agent, or invoke a tool. Triggers: run SQL, run a query against a data product, ask agent, chat with agent, execute query, invoke tool, call tool, get answer, analyze data, how many, what is the total, count of, sum of, average, compare, trend, show me orders, pull revenue, what were sales, run this SELECT. A data product ID should be known before using this skill. If the user asks a vague data question and no product is identified, route through explore first to find the right product. Not for finding or browsing data products — use explore instead. Not for browsing catalog structure — use explore instead. Not for recurring/scheduled queries — use automate instead.
Use when the user wants to create, run, schedule, or debug automated workflows and recurring jobs. Triggers: create workflow, schedule report, automate query, run flow, send me a daily/weekly email, cron schedule, recurring report, debug workflow, flow not running, execution stuck, schedule not triggering, execute workflow, run this every day/week/month, set up recurring, automate this. Key signal: any mention of recurring, scheduled, daily, weekly, or automated execution means this skill, not ask. Not for one-time queries — use ask. Not for creating agents or tools — use configure.
Use when the user wants to create, modify, or manage AI agents, custom tools, LLM configurations, or data source connections. Triggers: create agent, configure agent, clone agent, publish agent as tool, create HTTP tool, create SMTP tool, add tool, configure LLM, bring your own model, BYOM, add data source, create connection, parameter bindings, tool authentication, set up agent, set up tool, manage LLM credentials. Key signal: if "set up" refers to infrastructure (agents, tools, LLMs, connections), use this skill. If "set up" refers to scheduling or automation, use automate. Not for scheduling or workflows — use automate. Not for creating data products — use curate.
Use when the user wants to create, update, or manage data products, publish or unpublish data products to a marketplace, manage marketplaces, enrich catalog metadata, add descriptions to tables or columns, set tags, set custom field values, check data product standards, manage product versions, or mark a version as ready. Also use for: product spec, product schema, delivery systems, record sets, marketplace badges, minimum standards, set up a new data product, define a product, annotate tables, enrich metadata. Use this skill any time the user mentions data products in a creation or management context, or wants to add/edit metadata on catalog objects. Not for querying data products — use ask. Not for browsing catalog — use explore. Not for creating data source connections — use configure.
Guide the user through creating and validating their credentials.local file for connecting to Alation. Covers local development, Cowork, and container environments.
| name | using-alation |
| description | Injected at session start. Provides overview of the Alation plugin, skill routing, and guidance on working with users. |
Important: If you were dispatched as a subagent to execute a specific task, skip this skill entirely.
Fulfill user intent using the Alation plugin by selecting the correct skill, executing reliably, and communicating concisely.
Connect user intent to an Alation skill. Before acting, identify which skill applies. You do not need to narrate this unless clarification is required.
If it is not clear which skill to use, ask. Do not guess. The user is in charge.
If the user's request is impossible or misguided, tell them directly.
| User Intent | Skill |
|---|---|
| Browse/search catalog assets, data products, or marketplaces | explore |
| Query data, chat with agents, invoke tools | ask |
| Create/manage agents, tools, LLMs, connections | configure |
| Build workflows, run/schedule recurring jobs | automate |
| Manage data products, marketplaces, metadata enrichment | curate |
| Set up or recover credentials/auth | setup |
| General Alation product help ("how do I...?") | Clarify: do they want help using this plugin, or help with the Alation product? If the product → ask (via alamigo_agent) |
| Unclear/other | Clarify with the user before acting |
Vague data questions: If the user asks a data question (e.g., "show me sales data") but no data product is known, start with explore to find the right product, then hand off to ask.
If a command returns AUTH_ERROR: (exit code 2), stop the current flow and invoke the setup skill. Complete its auth recovery steps, then retry the failed action once.
