Skip to main content
Glama

google_ads_search_terms_review

Evaluates Google Ads search terms with six rules, sorting them into add, exclude, and watch buckets. Read-only review to improve campaigns without making changes.

Instructions

Score every search term in a Google Ads campaign against six hardcoded rules and split them into add / exclude / watch buckets. Returns {campaign_id, ad_group_id, period, target_cpa, target_cpa_source, add_candidates, exclude_candidates, watch_candidates, summary:{total_search_terms, add_count, exclude_count, watch_count}, intent_analysis?}. Each candidate has {search_term, action, match_type ('EXACT'|'PHRASE'), score (40-90), reason, metrics:{conversions, clicks, cost, ctr}}. target_cpa is resolved from the explicit argument first, then the campaign's bidding strategy, then last-30-days actual CPA; target_cpa_source reports which path ('explicit'|'bidding_strategy'|'actual'|'none'). New terms absent from the previous period are routed to watch_candidates. Read-only — emits candidates but does not add or exclude anything. Default period is LAST_7_DAYS. For keyword/N-gram overlap stats use google_ads_search_terms_analyze; for the raw query log use google_ads_search_terms_report.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
periodNoReporting window for the metrics. Default 'LAST_7_DAYS' — this tool is tuned for short-horizon comparison. Use LAST_14_DAYS or LAST_30_DAYS for longer baselines. This tool also reads the equal-length window immediately before the one you request, so only fixed-length windows are accepted — calendar constants such as THIS_MONTH are rejected rather than silently replaced. Also accepts an explicit range in GAQL spelling — "BETWEEN 'YYYY-MM-DD' AND 'YYYY-MM-DD'", both endpoints inclusive, in the account's time zone — for a window no trailing constant can reach (e.g. a single past calendar month). One asymmetry to know about: every constant except LAST_90_DAYS is resolved by Google Ads in the account's reporting time zone, whereas LAST_90_DAYS has no API constant and is expanded by mureo into the 90 days ending yesterday **on the server's date**, so its edges can differ by a day when the server and the account are in different zones. Pass an explicit range when the exact boundary matters.
target_cpaNoOptional explicit target CPA in account currency (e.g. 3000 = ¥3,000). Exclusion rule 4 fires at cost >= target_cpa * 2. Falls back to the campaign's bidding strategy target, then last-30-days actual CPA; if none can be resolved, CPA-gated rules are skipped.
campaign_idYesCampaign ID as a numeric string without dashes (e.g. '23743184133'). Obtain via google_ads_campaigns_list.
customer_idNoGoogle Ads customer ID as a 10-digit string without dashes (e.g. '1234567890'). Optional — falls back to GOOGLE_ADS_CUSTOMER_ID / GOOGLE_ADS_LOGIN_CUSTOMER_ID from the configured credentials when omitted.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed3 schema fields changedv0.17.1
    • addedInput schema / properties / period / anyOf
      Added value: +[
      +  {
      +    "enum": [
      +      "LAST_7_DAYS",
      +      "LAST_14_DAYS",
      +      "LAST_30_DAYS",
      +      "LAST_90_DAYS"
      +    ],
      +    "type": "string"
      +  },
      +  {
      +    "pattern": "\\ABETWEEN '([0-9]{4}-[0-9]{2}-[0-9]{2})' AND '([0-9]{4}-[0-9]{2}-[0-9]{2})'\\Z",
      +    "type": "string"
      +  }
      +]
    • changedInput schema / properties / period / description
      Previous value: -"Reporting window for the metrics. Default 'LAST_7_DAYS' — this tool is tuned for short-horizon comparison. Use LAST_14_DAYS or LAST_30_DAYS for longer baselines."New value: +"Reporting window for the metrics. Default 'LAST_7_DAYS' — this tool is tuned for short-horizon comparison. Use LAST_14_DAYS or LAST_30_DAYS for longer baselines. This tool also reads the equal-length window immediately before the one you request, so only fixed-length windows are accepted — calendar constants such as THIS_MONTH are rejected rather than silently replaced. Also accepts an explicit range in GAQL spelling — \"BETWEEN 'YYYY-MM-DD' AND 'YYYY-MM-DD'\", both endpoints inclusive, in the account's time zone — for a window no trailing constant can reach (e.g. a single past calendar month). One asymmetry to know about: every constant except LAST_90_DAYS is resolved by Google Ads in the account's reporting time zone, whereas LAST_90_DAYS has no API constant and is expanded by mureo into the 90 days ending yesterday **on the server's date**, so its edges can differ by a day when the server and the account are in different zones. Pass an explicit range when the exact boundary matters."
    • removedInput schema / properties / period / enum
      Removed value: -[
      -  "TODAY",
      -  "YESTERDAY",
      -  "LAST_7_DAYS",
      -  "LAST_14_DAYS",
      -  "LAST_30_DAYS",
      -  "LAST_90_DAYS",
      -  "THIS_MONTH",
      -  "LAST_MONTH"
      -]
  2. Changed1 schema field changedv0.10.37
    • addedInput schema / additionalProperties
      Added value: +false
  3. Addedv0.10.11
  4. Removedv0.10.9
  5. Addedv0.9.12
  6. Removedv0.9.6
  7. Addedv0.9.2
  8. Removedv0.9.1
  9. Addedv1.0.5

TDQS

A4.7/5.0
Behavior5/5

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

There are no annotations, so the description carries the full burden of behavioral disclosure. It explicitly states 'Read-only — emits candidates but does not add or exclude anything,' explains the target_cpa resolution precedence, and notes that new terms absent from the previous period are routed to watch_candidates. This is rich, useful behavioral context.

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

Conciseness5/5

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

The description is dense but well organized: main action, return object shape, candidate shape, fallback logic, read-only behavior, default period, and sibling alternatives. Each sentence earns its place, and the most important facts are front-loaded.

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?

Because there is no output schema, the description supplies a detailed return shape, candidate metrics, and the optional intent_analysis key. It also covers default period, CPA fallback, and the read-only guarantee, so an agent has everything needed to invoke and interpret the tool correctly.

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

Parameters3/5

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

Schema description coverage is 100%, so the baseline is 3. The JSON Schema already documents period, target_cpa, campaign_id, and customer_id thoroughly. The free-text description adds helpful context such as default period and fallback ordering, but the schema already carries the essential parameter semantics.

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+resource: 'Score every search term in a Google Ads campaign against six hardcoded rules and split them into add / exclude / watch buckets.' It clearly distinguishes this tool from the sibling analysis tools by naming google_ads_search_terms_analyze and google_ads_search_terms_report and giving each a separate use case.

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

Usage Guidelines5/5

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

The description explicitly routes the agent: 'For keyword/N-gram overlap stats use google_ads_search_terms_analyze; for the raw query log use google_ads_search_terms_report.' Combined with the stated default period and read-only behavior, the agent knows both when to choose this tool and when to select an alternative.

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

Install Server

Other Tools