Add ready-made skills from the skill hub to your workspace so conversational AI can produce catalog health reports, governance audits, lineage analysis, AI usage reports, and more.
Add skills from skill hub
The skill hub is a library of ready-made skills that Atlan maintains. Each one teaches conversational AI how to handle a specific kind of request, such as auditing governance coverage or reporting on AI usage, with the steps, queries, and output format already worked out. Add a skill from the hub instead of writing one from scratch, then edit it to fit how your team works.
Prerequisites
Before you begin, make sure you have:
- Set up conversational AI for your tenant. See Set up conversational AI. The New skill option appears only when conversational AI is enabled.
- Permission to create skills in your workspace.
Add skill
- From the navigation menu on any screen in Atlan, click New, then click New skill.
- Click Start from templates. The Skill hub opens with the skills available to add.
- Find the skill you want, and click Add skill on its card.
Atlan adds a copy of the skill to your workspace and opens its profile. From there you can review its instructions and edit them. Your copy belongs to your workspace: changes you make don't affect the hub, and conversational AI uses the skill in your next conversation.
To use a skill, ask conversational AI for what it does, for example, "Give me a governance coverage report" or "What breaks downstream if I change this table?" Skills are applied in Deep and Adaptive modes. See Conversational AI capabilities.
Available skills
The hub grows over time, so the list in your workspace may include skills that aren't described here.
Reports and assessments
| Skill | What it produces |
|---|---|
data-catalog-health-report | A health assessment of your catalog: asset coverage, ownership, documentation, glossary adoption, certification rates, and recommendations. |
governance-coverage-report | A governance audit: ownership gaps, certification rates, glossary adoption, documentation coverage, and prioritized remediation actions. |
data-landscape-executive-overview | An executive-level summary of your catalog: scale, sources, organization, key metrics, and observations. |
data-domain-readiness-assessment | A production-readiness review of one data domain: governance maturity, asset inventory, ownership, certification, glossary coverage, and a go/no-go recommendation. |
asset-deep-dive-report | A full profile of a single asset: metadata, schema, ownership, business context, lineage, documentation, and quality signals. |
ai-usage-analysis-report | An analysis of AI usage across conversational AI, MCP, and Slack: adoption, tool calls, query patterns, errors, latency, and gaps. |
account-ai-report | An AI usage report for one tenant across conversational AI, MCP, Slack, and Teams, with usage trends, user stories, failures, and recommendations. Add --days N to change the 30-day lookback. |
The report skills publish their output as an artifact that you can preview, revise, and download.
Step-by-step recipes
| Skill | What it does |
|---|---|
lineage-impact-analysis | Assesses downstream impact before you change a table, column, or model. Use it to answer "What breaks if I change this?" |
metadata-completeness-audit | Finds the worst-documented assets, recently modified assets missing documentation, and stale certifications. |
user-onboarding | Helps a first-time user get a useful answer quickly: asks about their role and suggests two or three prompts to try. Useful for "what can you do?" and "how do I get started?" |
Query patterns
| Skill | What it does |
|---|---|
lineage-coverage-analytics | Calculates lineage coverage overall and by connector and asset type, and finds assets without lineage. |
usage-analytics | Answers product-usage questions such as active users, feature adoption, retention, and stickiness. The AI usage report skills build on it, so add it alongside them. |
See also
- Artifacts in conversational AI: create your own skill from chat.
- Custom instructions for conversational AI: when to use a skill instead of a custom instruction.