Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@tableau-graphql-mcp Which workbooks break if I remove the SALES column?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
tableau-graphql-mcp turns your Tableau site's Metadata API into a set of MCP tools, so an AI assistant (Claude, Cursor, Cline, and others) can answer lineage questions in plain language:
"If I drop the column SALES, which workbooks break?"
"What tables does the Sales Overview workbook depend on?"
"Which calculated fields reference Profit, and on which dashboards?"
"Who should I notify before changing the DIM_CUSTOMER table?"
It ships seven curated tools: a universal GraphQL passthrough, live schema introspection, an embedded library of correct query templates, a robust where_used resolver, a multi-hop impact_analysis, a substring content search, and a connection probe. Together they let the model answer any lineage question, not just a fixed menu.
Any question, done right. graphql_query runs any read-only GraphQL; introspect_schema and a built-in cheat-sheet plus 28 worked examples keep the model's queries correct.
True impact analysis (multi-hop). impact_analysis follows the whole dependency chain (a calc built on a calc built on a column is included) and returns the full blast radius plus the de-duplicated owners to notify, not just direct references.
Works everywhere. Tableau Server and Cloud. The REST API version and the GraphQL endpoint (/api/metadata/graphql, with a /relationship-service-war/graphql fallback) are auto-detected.
Robust lineage without Catalog. where_used resolves workbooks via core lineage (referencedByFields -> sheets -> workbook), so it works even when the Data Management add-on's downstreamWorkbooks is empty.
No silent truncation. graphql_query flags partial_results when a query hits the node limit, and search_content reports scanned/total coverage, so a truncated answer is never mistaken for a complete one.
Tiny and safe. Read-only, stdio-only (no inbound port), secrets from env only, and no dependencies beyond the MCP SDK (stdlib urllib for HTTP).
You need uv (curl -LsSf https://astral.sh/uv/install.sh | sh) and a Tableau Personal Access Token.
Claude Code, one line:
claude mcp add tableau-graphql \
-e TABLEAU_SERVER=https://10ax.online.tableau.com \
-e TABLEAU_SITE_CONTENT_URL=YourSite \
-e TABLEAU_PAT_NAME=my-token \
-e TABLEAU_PAT_SECRET=the-full-secret \
-- uvx tableau-graphql-mcpThat's all. uvx fetches the package from PyPI and runs it in an isolated environment; nothing to clone or install (and no git required). Then ask Claude a lineage question.
Related MCP server: MetaGraph-MCP
All configuration is via environment variables (set them in your client's env block, never on the command line).
Env var | Required | Default | Description |
| yes | n/a |
|
| no |
| Site slug (the part after |
| yes¹ | n/a | Personal Access Token name. |
| yes¹ | n/a | PAT secret: the whole string, do not split on |
| no |
| Per-request timeout (seconds). |
| no | auto | REST API version; else read from |
| no | auto | Override the GraphQL path; else auto-detected. |
| no | n/a | Advanced: a pre-obtained |
| no | n/a | Advanced: a browser session cookie (SSO fallback). |
¹ Provide a PAT (TABLEAU_PAT_NAME + TABLEAU_PAT_SECRET) or an advanced TABLEAU_AUTH_TOKEN / TABLEAU_COOKIE.
Edit claude_desktop_config.json (macOS: ~/Library/Application Support/Claude/, Windows: %APPDATA%\Claude\):
{
"mcpServers": {
"tableau-graphql": {
"command": "uvx",
"args": ["tableau-graphql-mcp"],
"env": {
"TABLEAU_SERVER": "https://10ax.online.tableau.com",
"TABLEAU_SITE_CONTENT_URL": "YourSite",
"TABLEAU_PAT_NAME": "my-token",
"TABLEAU_PAT_SECRET": "the-full-secret"
}
}
}
}Fully quit and reopen Claude Desktop, then check the tools menu.
Every client uses the same mcpServers schema shown above. Add the same block to:
Cursor: ~/.cursor/mcp.json (global) or .cursor/mcp.json (project).
Cline: the MCP Servers panel, then Configure, into cline_mcp_settings.json.
Windsurf: ~/.codeium/windsurf/mcp_config.json.
On Windows, if uvx isn't found by the GUI app, use its absolute path (e.g. %USERPROFILE%\.local\bin\uvx.exe).
Tool | What it does | Key args |
| Run any read-only Metadata API GraphQL query. The general tool for any lineage question. |
|
| Live schema introspection: list entry points, or a type's exact fields. |
|
| A schema cheat-sheet plus 28 curated question-to-GraphQL templates (8 categories). |
|
| Which workbooks/datasources use given column / field / table names (robust one-hop core-lineage resolution). |
|
| Full transitive multi-hop blast radius of a column/field/table: every dependent field, plus affected sheets, dashboards, workbooks, and owners to notify. |
|
| Find content whose name contains a term (case-insensitive substring), across workbooks, datasources, tables (and optionally fields/columns), with coverage numbers. |
|
| Connected server, site, versions, endpoint, auth, and whether Catalog lineage is available. | none |
All tools are read-only. The Metadata API has no mutations.
Once connected, try:
"Use server_info to confirm what you're connected to."
"Search for anything with 'revenue' in the name."
"Which workbooks use the columns SALES, PROFIT and DISCOUNT? Group by owner."
"Show me the field-to-source-column map for the 'Sales Overview' workbook."
"List every calculated field in that workbook with its formula."
"Which published datasources feed workbooks in the Analytics project, and which are uncertified?"
"Run impact_analysis on the 'Profit Ratio' field: every dependent sheet, dashboard, workbook, and owner to notify."
"What is the blast radius of dropping the DIM_CUSTOMER table: workbooks, sheets, and owners to notify?"
The server speaks MCP over stdio to the client and HTTPS to Tableau: it signs in with your PAT to get an X-Tableau-Auth token (auto-refreshed on expiry), auto-detects the REST API version and the GraphQL endpoint, then forwards queries to the Metadata API. Nothing is stored; every answer is live.
Read-only, enforced. Only GraphQL queries: no writes, no shell. graphql_query rejects mutation/subscription operations, and the Metadata API is query-only regardless.
Local and stdio-only. No inbound network port is opened.
Secrets from env only. Never passed as tool arguments, never logged, never returned in output.
Least privilege. The PAT inherits your Tableau permissions; the API only returns content you can see.
Pin a version in production: uvx tableau-graphql-mcp==0.1.0.
See SECURITY.md.
Symptom | Fix |
Server doesn't appear | Fully quit and relaunch the client; check the config path and JSON validity. |
| Install |
Sign-in fails (401) | Check the PAT name/secret and |
"Could not reach the Metadata API" | On Tableau Server, an admin must enable it: |
Empty | Expected without the Data Management add-on; use the |
Inspect the server directly with the MCP Inspector:
npx @modelcontextprotocol/inspector uvx tableau-graphql-mcpgit clone https://github.com/tdries/tableau-graphQL-mcp && cd tableau-graphQL-mcp
uv sync --all-extras
uv run tableau-graphql-mcp # run from source
uv run pytest --cov=tableau_graphql_mcp # tests + coverage (offline; no Tableau needed)
uv run ruff check . # lint
uv run ruff format --check . # formatThe same three gates (lint, format, tests with a 85% coverage floor) run in CI across Linux/macOS/Windows and Python 3.10 to 3.13. Coverage is reported to Codecov and the code is scanned by CodeQL on every push.
The same gates run in CI (Linux/macOS/Windows, Python 3.10 to 3.13): ruff check,
ruff format --check, mypy --strict, and pytest with a coverage floor. The package
ships a PEP 561 py.typed marker, so importing it gives your type checker full types.
Contributions welcome: see CONTRIBUTING.md and the Code of Conduct.
Shipped: published on PyPI and listed on the official MCP registry. Next:
Optional Data Management path: richer downstreamWorkbooks when Catalog is present
More curated query templates in lineage_examples
Optional response caching for repeated introspection within a session
Ideas and votes welcome in Discussions.
Run ANY read-only GraphQL query against the Tableau Metadata API. This is the general-purpose tool; use it for any lineage question. Returns {"data": ..., "errors": ...}.
The Metadata API is a GraphQL graph of Tableau content (workbooks, sheets, dashboards,
datasources, fields) and, with the Data Management add-on, physical assets (databases,
tables, columns) and their upstream/downstream lineage.
How to write a correct query:
- Filters are EXACT and case-sensitive: `filter: {name: "X"}` or `filter: {nameWithin: ["X","Y"]}`
(nameWithin is the only multi/OR match; there is NO substring or regex).
- Every list field has a `<name>Connection` variant with `first`/`offset`/`after` + `pageInfo`
(page size max 1000). Keep one query under ~20,000 nodes; narrow filters and page.
- Fields is an interface; branch with `__typename` and inline fragments
(`... on ColumnField { columns { name } }`, `... on CalculatedField { formula }`).
- Reach a field's owning workbook via `datasource { ... on EmbeddedDatasource { workbook { name } } }`.
- `downstream*` fields (downstreamWorkbooks/Owners, external tables/columns) need Data Management
(Tableau Catalog) and are often empty otherwise; then resolve via core lineage
(referencedByFields -> sheets -> workbook), or just call the `where_used` tool.
Entry points include: workbooks, sheets, dashboards, publishedDatasources, embeddedDatasources,
fields, columnFields, calculatedFields, columns, databaseTables, customSQLTables, databases,
flows, tableauUsers, dataQualityWarnings. Call `lineage_examples` for ready-made query templates
and a schema cheat-sheet, or `introspect_schema` to inspect any type's exact fields.
Read-only: mutation and subscription operations are rejected. If a query exceeds the
~20,000-node limit, the response is flagged with `partial_results: true` and a `warning`
(it does NOT auto-page an arbitrary query) so you never mistake a truncated result for a
complete one.
| Name | Required | Description | Default |
|---|---|---|---|
| query | Yes | ||
| variables | No |
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Discloses critical runtime behaviors: read-only enforcement, exact case-sensitive filters, page size max 1000, ~20k node limit, partial_results flag, no auto-paging, and Data Management dependency — all beyond the absent annotations.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Long but efficiently structured with front-loaded purpose, then actionable guidelines and caveats. Every paragraph earns its place, and the organization makes the density navigable.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Covers return format, error handling, partial results, Data Management prerequisites, entry points, and helper tools, making it fully self-contained for a complex API with no output schema shown.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Despite 0% schema coverage, the description compensates with detailed GraphQL query construction guidance (filters, paging, inline fragments, connections). However, the optional 'variables' parameter is not explicitly explained, leaving a minor gap.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
Clearly states 'Run ANY read-only GraphQL query against the Tableau Metadata API' and identifies itself as the general-purpose lineage tool, distinguishing from specialized siblings like where_used and lineage_examples.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Explicitly directs when to use it ('use it for any lineage question'), when to call where_used for downstream lineage without Data Management, and points to lineage_examples and introspect_schema for templates and schema inspection.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Full transitive (MULTI-HOP) downstream impact of a column, field, or table. Returns every field that directly OR indirectly depends on it, and all affected sheets, dashboards, workbooks, plus the de-duplicated set of OWNERS to notify before a change.
This is the "what breaks if I change/drop this?" tool. Unlike `where_used` (one core-lineage
hop), it follows the whole dependency chain — a calc built on a calc built on the column is
included. `name` is exact and case-sensitive. Workbooks are derived from downstream
sheets/dashboards (the flat downstreamWorkbooks edge is unreliable). If it returns nothing on a
site without Data Management, the transitive lineage may not be indexed there — use `where_used`
for the direct references.
| Name | Required | Description | Default |
|---|---|---|---|
| name | Yes |
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, but description discloses exact case-sensitive matching, transitive multi-hop behavior, derivation of workbooks from downstream sheets/dashboards, and site-level Data Management limitation. It also notes the flat downstreamWorkbooks edge is unreliable, which is valuable behavior beyond schema.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is longer than typical but every sentence adds unique operational context: main behavior, comparison, matching rules, workbook derivation, and fallback. Front-loaded with the core purpose and no filler.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a one-parameter tool with no annotations and minimal schema, the description covers output contents, owner de-duplication, caveats about Data Management, and relationship to sibling tool. It provides enough context for correct invocation without needing additional documentation.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Input schema only has 'name' with no description. Description adds that `name` is exact and case-sensitive and identifies the target as a column, field, or table, giving the parameter actionable meaning.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
Description states 'Full transitive (MULTI-HOP) downstream impact' and clarifies 'what breaks if I change/drop this?' It specifically distinguishes from `where_used`, making sibling differentiation explicit. This is a strong purpose statement.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Explicitly frames when to use ('what breaks if I change/drop this?') and contrasts with `where_used` ('one core-lineage hop'). Also provides a fallback direction when transitive lineage may not be indexed ('use `where_used` for the direct references').
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Introspect the live Metadata API GraphQL schema (introspection is enabled).
With no argument: returns every Query entry point (with its args) and the full list of
type names. With `type_name` (e.g. "Column", "Workbook", "DatabaseTable", "CalculatedField"):
returns that type's fields, their result types, and args, so you can write a correct query
against exactly what this server exposes.
| Name | Required | Description | Default |
|---|---|---|---|
| type_name | No |
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden of behavioral disclosure. It explicitly states that introspection is enabled, and details exactly what is returned in each mode (Query entry points with args, full type list, or type fields with result types and args). This gives the agent a precise understanding of the tool's behavior.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is compact and front-loaded, with the core action in the first sentence. The second sentence efficiently covers both invocation modes and ties the purpose to writing correct queries. Every clause earns its place; there is no fluff or repetition.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's simplicity (one optional parameter) and the presence of an output schema, the description is complete. It covers both usage scenarios, explains the data returned, and provides examples of type names. No important details are missing for an agent to invoke it correctly.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The input schema provides only the parameter name 'type_name' with no description (0% coverage). The description compensates fully by explaining the no-argument default, what happens when type_name is provided, and giving concrete examples of valid values. This adds substantial meaning beyond the bare schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description opens with a specific verb and resource: 'Introspect the live Metadata API GraphQL schema', which clearly distinguishes this from sibling tools like graphql_query. It then elaborates with two concrete usage modes (no argument vs. with type_name), making the purpose unmistakable.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides clear context on when to use the tool: before writing a query, to understand exactly what the server exposes. It explains both invocation patterns but does not explicitly name alternative tools or state when not to use it, so it stops short of a 5.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Return a schema cheat-sheet and a library of curated lineage questions with their
correct GraphQL queries (+ example variables). Read this before composing a graphql_query.
Categories: impact, provenance, calc, datasource, search, inventory, governance, ownership.
Pass one to filter; omit to get them all. Each example has: question, graphql, variables, notes.
| Name | Required | Description | Default |
|---|---|---|---|
| category | No |
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the burden. It discloses the parameter behavior (pass one to filter, omit to get all) and the structure of each example (question, graphql, variables, notes). It also indicates the categories available, which adds behavioral context beyond the bare schema. There is no contradiction with annotations since none exist.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is compact and well-structured, with the main purpose in the first sentence, followed by category list and example structure. Every sentence adds value, and there is no redundant information.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool has only one optional parameter and an output schema exists (which defines the returned structure), the description covers all essential aspects: what it returns, how to filter, and what each example contains. It is sufficiently complete for an agent to invoke correctly without further clarification.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0%, so the description must compensate for the category parameter. It does so thoroughly by listing all valid categories (impact, provenance, calc, datasource, search, inventory, governance, ownership) and explaining that passing one filters while omitting returns all. This fully explains both the meaning and usage of the parameter.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses a clear verb 'Return' and specifies the resource: a schema cheat-sheet and a curated library of lineage questions with GraphQL queries and variables. It distinguishes itself from sibling tools like graphql_query, which executes queries, and where_used/impact_analysis, which are specific lineage analysis tools.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description explicitly says 'Read this before composing a `graphql_query`', providing clear when-to-use guidance. It also explains the optional category filter and the ability to omit it, giving clear usage context. However, it does not explicitly mention when not to use it or contrast with other siblings beyond graphql_query.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Find content whose NAME contains term (case-insensitive SUBSTRING). Use this when you
only know part of a name, since every other tool and the Metadata API filter are exact-match.
Searches workbooks, published datasources, and database tables by default. Pass `types` to
choose from: "workbook", "datasource", "table", "field", "column". Returns matches grouped by
type. It pages through content client-side, so on a very large site it scans the first ~1200
of each type and says so in `note`; once you know the exact name, prefer `graphql_query` or `where_used`.
| Name | Required | Description | Default |
|---|---|---|---|
| term | Yes | ||
| types | No |
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden of behavioral disclosure. It reveals default search scope (workbooks, published datasources, database tables), optional `types` filtering, grouping of matches by type, client-side pagination with a ~1200 item scanning limit per type, and the presence of a `note` field indicating this limitation. This goes beyond the schema and makes hidden behaviors explicit.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is compact yet information-dense, front-loading the core purpose in the first sentence. Every subsequent sentence adds essential detail—usage context, defaults, side effects, and alternatives—without redundancy or padding.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the output schema exists (so return structure need not be spelled out), the description covers all invocation-relevant aspects: what it searches, how to narrow results, how results are organized, pagination limitations, and when to choose an alternative. It is complete for an agent to select and call the tool correctly.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0%, so the description must compensate. It fully explains both parameters: `term` as a case-insensitive substring matched against the NAME field, and `types` with the allowed values (workbook, datasource, table, field, column) and default behavior when not passed. This adds all necessary meaning beyond the bare schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses a specific verb ('Find') and resource ('content whose NAME contains `term`') and clarifies it performs a case-insensitive substring match. It also distinguishes this tool from siblings by stating that every other tool and the Metadata API filter are exact-match, making its purpose and scope immediately clear.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description explicitly states when to use this tool ('Use this when you only know part of a name') and contrasts it with alternatives, saying to prefer `graphql_query` or `where_used` once the exact name is known. This provides clear decision guidance for the agent.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Report the connected Tableau environment: server URL, site, product & REST API version, which Metadata API endpoint is in use, the auth method, and whether external-asset (Data Management / Catalog) lineage appears available. Good first call to confirm the connection and understand what lineage depth to expect.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the burden of behavioral disclosure. It reveals the tool reports environment information and hints at read-only nature through the verb 'report,' but it does not explicitly state that it makes no changes, has no side effects, or specify any other behavioral characteristics. The description adds useful context about lineage availability but lacks definitive safety statements.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is two sentences, front-loaded with the purpose ('Report the connected Tableau environment') followed by the specific data points and a clear usage recommendation. Every sentence earns its place with no redundancy.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a zero-parameter tool with an output schema (even if not shown), the description is complete. It specifies what information is returned, the intended first-call use case, and what the user can infer about lineage depth. No additional context is needed for this simple tool.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The input schema is empty (0 parameters), giving a baseline of 4 per the rubric. The description does not mention parameters because there are none, which is appropriate. No additional parameter semantics are needed.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool reports the connected Tableau environment, listing specific details (server URL, site, versions, auth method, lineage availability). This specific verb+resource clearly distinguishes it from sibling tools like graphql_query or where_used, which serve different purposes.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Explicitly recommends this as a 'good first call to confirm the connection and understand what lineage depth to expect,' which provides clear context for when to use the tool. However, it does not explicitly mention when not to use it or name alternatives, so it stops short of a full 5.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Find which workbooks (and published datasources) USE the given names, the common 'where is this used / impact analysis' question, resolved robustly.
`names` are EXACT, case-sensitive names of any of: a Snowflake/DB column, a Tableau
field or alias, or a database table. Pass several to check them in one call. Results
group by workbook, showing how each matched (column / field / whole table, with schema
and source table) and which worksheets use it.
This uses CORE lineage (referencedByFields -> field.sheets -> workbook, and
field.datasource -> workbook), so it works even without the Data Management add-on where
`downstreamWorkbooks` is empty. For other shapes of question, use `graphql_query`.
| Name | Required | Description | Default |
|---|---|---|---|
| names | Yes |
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden and excels: it discloses exact case-sensitive matching, grouping of results by workbook, and the underlying core lineage mechanism, setting clear expectations for the agent.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise yet dense, with each sentence serving a distinct purpose: purpose, parameter details, result behavior, and alternative tool. There is no waste or redundancy, and important details are front-loaded.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given that an output schema exists, the description doesn't need to explain return values. It fully covers the use case, parameter semantics, behavioral nuances, and alternatives, making it complete for a query tool with minimal parameters.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0%, but the description thoroughly explains the `names` parameter: what types of names are accepted (column, field, alias, table), exact and case-sensitive matching, and support for multiple names in one call. This fully compensates for the lack of schema-level parameter descriptions.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool finds workbooks and published datasources that use given names, with a specific verb and resource. It explicitly names graphql_query as an alternative for other question shapes, distinguishing from at least that sibling.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
It provides explicit context for when to use this tool for 'where is this used / impact analysis' questions and says to use graphql_query for other shapes. It also explains that it works without the Data Management add-on, giving clear usage exclusions and conditions.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections. Dates show when Glama detected each change.
v0.1.0graphql_queryimpact_analysisintrospect_schemalineage_examplessearch_contentserver_infowhere_usedEach tool has a clearly distinct role: general-purpose GraphQL querying, schema introspection, example retrieval, direct lineage lookup, transitive impact analysis, fuzzy name search, and environment info. The overlap between where_used and impact_analysis is explicitly explained, eliminating ambiguity.
Tool names are all lowercase snake_case but mix verb-first (introspect_schema, search_content), noun-first (graphql_query, lineage_examples, impact_analysis, server_info), and the unconventional where_used. There is no consistent verb_noun pattern, making the naming conventions mixed but still readable.
Seven tools is a well-scoped number for a Tableau Metadata API server. It includes a general-purpose query tool plus specialized helpers that earn their place, neither sparse nor overwhelming.
The tool set comprehensively covers the domain: raw GraphQL access, schema inspection, example templates, direct and transitive lineage queries, fuzzy search, and environment diagnostics. graphql_query fills any niche gaps, leaving no obvious dead ends.
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
Governed data discovery, exact queries, decisions, simulations, and runtime utilities over MCP.
Query your org's data in natural language — read-only MCP access to SQL, NoSQL, files & warehouses.
Governed BI MCP. Ask questions of live company data and list workspace sources via OAuth.
Query your warehouse or a CSV with Claude/ChatGPT over MCP, governed by table-level ACL + audit.
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/tdries/tableau-graphQL-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server
