Skip to main content
skill CX Codex

knowledge-catalog-discovery — Claude Code skill: what it does

A set of discovery scripts to search data assets and inspect metadata in Knowledge Catalog through Claude Code.

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add cline/skills --skill knowledge-catalog-discovery -a claude-code
Open on GitHub

What knowledge-catalog-discovery does

The skill lets coding assistants search and inspect data assets in Google Cloud Dataplex Knowledge Catalog. A set of Node.js scripts finds tables, datasets, or views, pulling back schema details and relationships between assets.

Four scripts handle discovery: search_entries supports Dataplex query syntax with grouping and boolean operators, lookup_entry inspects a single entry across different view modes, lookup_context retrieves metadata for up to 10 resources in the same location, and search_aspect_types queries catalog aspects. Executing these tools requires Node.js in your environment.

Environment variables load automatically from .env files without manual user prompting. Broad searches without resource filters run slowly and consume significant capacity, so exploratory queries need a defined pageSize.

What it helps with

  • find tables and views in Dataplex using query filters
  • inspect metadata and aspect types for a specific catalog entry
  • retrieve relationships for multiple assets within one location
  • search available catalog aspect types

Example prompts for the agent

  • Search for tables matching orders in project analytics
  • Retrieve full metadata and aspects for the customer dataset in location us
  • Find catalog aspect types related to data quality

How to install knowledge-catalog-discovery

1Skills CLI (Vercel)

npx skills add cline/skills --skill knowledge-catalog-discovery -a claude-code

2Manually: put the files in place

~/.claude/skills/knowledge-catalog-discovery/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/knowledge-catalog-discovery/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about knowledge-catalog-discovery

How do I install knowledge-catalog-discovery in Claude Code?

Run npx skills add cline/skills --skill knowledge-catalog-discovery -a claude-code in your terminal. This saves the skill files into ~/.claude/skills/knowledge-catalog-discovery/SKILL.md.

What does knowledge-catalog-discovery do?

It enables your agent to search assets like tables and views in Knowledge Catalog, inspect entries, and explore metadata aspect types.

What are the runtime requirements for these scripts?

You need Node.js installed to execute the JavaScript files. Credentials and settings are loaded automatically from project .env files.

Source text

SKILL.md · Apache-2.0
---
name: knowledge-catalog-discovery
description: Use these skills when you need to discover and explore data assets in the Knowledge Catalog. It allows you to search for entries, lookup specific metadata, and explore aspect types to understand your data platform's assets.
---

## Usage

All scripts can be executed using Node.js. Replace `<param_name>` and `<param_value>` with actual values.

**Bash:**
`node <skill_dir>/scripts/<script_name>.js '{"<param_name>": "<param_value>"}'`

**PowerShell:**
`node <skill_dir>/scripts/<script_name>.js '{\"<param_name>\": \"<param_value>\"}'`

Note: The scripts automatically load the environment variables from various .env files. Do not ask the user to set vars unless skill executions fails due to env var absence.


## Scripts


### lookup_context

Retrieves rich metadata regarding one or more data assets along with their relationships.

#### Parameters

| Name | Type | Description | Required | Default |
| :--- | :--- | :--- | :--- | :--- |
| resources | array | Required. A list of up to 10 resource names. Resources may belong to different projects, but all must belong to the same location. | Yes |  |


---

### lookup_entry

Retrieves a specific metadata regarding a data asset (e.g. table/dataset/view) from Catalog

#### Parameters

| Name | Type | Description | Required | Default |
| :--- | :--- | :--- | :--- | :--- |
| entry | string | Required. The resource name of the Entry in the following form: projects/{project}/locations/{location}/entryGroups/{entryGroup}/entries/{entry}. | Yes |  |
| view | integer | 
				## Argument: view

				**Type:** Integer

				**Description:** Optional. Specifies the parts of the entry and its aspects to return.

				**Possible Values:**

				*   1 (BASIC): Returns entry without aspects.
				*   2 (FULL): Return all required aspects and the keys of non-required aspects. (Default)
				*   3 (CUSTOM): Return the entry and aspects requested in aspect_types field (at most 100 aspects). Always use this view when aspect_types is not empty.
				*   4 (ALL): Return the entry and both required and optional aspects (at most 100 aspects)
				 | No | `2` |
| aspectTypes | array | Optional. Limits the aspects returned to the provided aspect types. It only works when used together with CUSTOM view. | No | `[]` |


---

### search_aspect_types

Search aspect types relevant to the query.

#### Parameters

| Name | Type | Description | Required | Default |
| :--- | :--- | :--- | :--- | :--- |
| query | string | The query against which aspect type should be matched. | Yes |  |
| pageSize | integer | Number of returned aspect types in the search page. | No | `5` |
| orderBy | string | Specifies the ordering of results. Supported values are: relevance, last_modified_timestamp, last_modified_timestamp asc | No | `relevance` |


---

### search_entries

Searches for data assets (eg. table/dataset/view) in Catalog based on the provided search query.

#### Parameters

| Name | Type | Description | Required | Default |
| :--- | :--- | :--- | :--- | :--- |
| query | string | A query string for searching entries, following Dataplex search syntax. Supports logical operators (AND, OR, NOT) and grouping. For example, to find a table that might have been renamed, you could use 'type:table (name:books OR fiction)'. This can be more efficient than multiple separate calls.Warning: Performing broad searches without specific filters (e.g., type:table) can be slow and consume significant resources. When performing exploratory searches, always use the pageSize parameter to limit the number of results returned. | Yes |  |
| scope | string | A scope limits the search space to a particular project or organization. It must be in the format: organizations/<org_id> or projects/<project_id> or projects/<project_number>. | No | `` |
| pageSize | integer | Number of results in the search page. | No | `5` |
| orderBy | string | Specifies the ordering of results. Supported values are: relevance, last_modified_timestamp, last_modified_timestamp asc | No | `relevance` |


---

© cline · Apache-2.0 · cline/skills

Similar skills

All skills for Data and databases →

From the same set17

frontend-design The skill makes the agent pick a bold visual direction before writing code, then hold it through every detail. Useful when a generated interface should not look generated. CX Codex +1 7,300installs review-team Runs several independent reviewers against the same change. Useful when you want more than one perspective instead of a single pass. CX Codex +1 3,300installs desktop-commander-overview Tells the agent what the Desktop Commander MCP can do: persistent sessions, long-running processes, files outside the workspace, Excel and PDF, SSH. CX Codex +1 2,500installs site-specification Turns brief website ideas or brand images into a JSON design specification with layout requirements, target audience, and Google Fonts. CX Codex +1 2,100installs data-analyst Turns your agent into an analyst: it pins down the question, runs a bounded ClickHouse query, and explains what the number actually means. CX Codex +1 2,200installs vibe-prospecting Turns your agent into a prospecting tool: company and contact data, roles, tech stack, business events, site changes. Handy when you need a lead list. CX Codex +1 1,400installs searching-sourcegraph Tells the agent which Sourcegraph MCP tool to reach for and in what order. Useful when you need to find an implementation or make sense of an unfamiliar system. CX Codex +1 1,400installs endor-setup This skill gets endorctl ready to work: checks the install, downloads the binary, handles authentication and runs a vulnerability scan. CX Codex +1 1,200installs

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor