ADM Google Ads

List keywords with performance

get_keywords

Read-onlyIdempotent

Lists positive keywords (including PAUSED, never REMOVED) with performance for the last N complete days. text uses Google notation: "phrase", [exact], broad as bare words; also match_type, status, quality_score (1-10, null if not enough data), max_cpc, campaign, ad_group, and impressions, clicks, cost, conversions, conversions_value, ctr, average_cpc. Filter with status (ENABLED or PAUSED), match_type (EXACT, PHRASE or BROAD) and text_contains (case-insensitive, ignores the quotes and brackets). compact=true returns only keyword id, text, status, campaign and ad group. Optionally limit to one campaign or one ad group. Sorted by cost descending (compact: by ad group, then id). Paginated via cursor. Negative keywords are in get_negative_keywords.

Input Schema

Name	Required	Description
`days`	No	Reporting window in whole days in the account's time zone, excluding today: 1 to 30. Default 7. Short windows hold little data, and conversions from the most recent days may not be reported yet.
`limit`	No	Page size, 1-200. Default 50.
`cursor`	No	Opaque next_cursor from the previous page. Keep the other arguments identical.
`status`	No	ENABLED or PAUSED. Omit to list both.
`compact`	No	When true, each keyword is only id, text, status, campaign and ad group.
`match_type`	No	EXACT, PHRASE or BROAD. Omit to list all match types.
`ad_group_id`	No	Optional ad group id (from get_ad_groups) to limit results to one ad group.
`campaign_id`	No	Optional campaign id (from get_campaigns) to limit results to one campaign.
`customer_id`	Yes	Google Ads customer ID from get_accounts, e.g. 123-456-7890.
`text_contains`	No	Case-insensitive substring of the keyword text.

Schema Changelog

Changes observed during successful MCP inspections.

Changed3 schema fields changedOct 12, 2026

addedInput schema / properties / ad_group_id

Added value: +{
+  "default": null,
+  "description": "Optional ad group id (from get_ad_groups) to limit results to one ad group.",
+  "type": [
+    "integer",
+    "null"
+  ]
+}

addedInput schema / properties / compact

Added value: +{
+  "default": null,
+  "description": "When true, each keyword is only id, text, status, campaign and ad group.",
+  "type": [
+    "boolean",
+    "null"
+  ]
+}

changedInput schema / properties / days / description

Previous value: -"Reporting window in whole days in the account's time zone, excluding today: 7, 14 or 30. Default 7."New value: +"Reporting window in whole days in the account's time zone, excluding today: 1 to 30. Default 7. Short windows hold little data, and conversions from the most recent days may not be reported yet."

First observedOct 9, 2026

TDQS

A4.4/5.0

Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare readOnlyHint, idempotentHint and openWorldHint=false, so safety is covered. The description adds genuinely non-structured behavior: REMOVED keywords are excluded, results are sorted by cost descending (ad group then id when compact), and paging is cursor-based. It could still say more about performance-data staleness bounds, but this is well beyond annotation coverage.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Effectively one dense paragraph but front-loaded with scope, then fields, then filters, then sorting/pagination and the sibling pointer. Every clause carries information, though the run-on delivery of the returned-field list is denser than it needs to be.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

With no output schema, the description carries the return-value burden and does so by enumerating the returned metric and dimension fields plus the compact field subset. Combined with the disclosed sort order, pagination, and filter semantics, an agent has everything needed to call this correctly for a 10-parameter tool.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3, but the description adds real meaning: the Google text notation ('phrase', [exact], bare = broad), the note that text_contains is case-insensitive and ignores quotes/brackets, and exactly which five fields compact=true returns. These are semantic clarifications the schema alone does not convey.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

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 plus a crucial scope qualifier: 'Lists positive keywords (including PAUSED, never REMOVED) with performance for the last N complete days.' It explicitly routes the disjoint sibling case at the end ('Negative keywords are in get_negative_keywords'), so an agent can distinguish it from the other ~20 siblings without opening schemas.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives clear routing for the main ambiguity (negative keywords live in a sibling tool) and states how to use compact=true versus full output. It stops short of explicitly contrasting with adjacent tools like get_search_terms, so it is strong context without a complete alternatives map.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All ADM Google Ads tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources

Need Help?