Google Ads (v25)

PREREQUISITE: Read ../_mureo-shared/SKILL.md for auth, global flags, and security rules.

Tool names below refer to the built-in platforms; a namespaced variant (see _mureo-shared → Namespaced built-in variants) is used with its prefix substituted.

There is no mureo google-ads … CLI command. Every operation below is an MCP tool (google_ads_*). Call the tool directly — never run or suggest a mureo google-ads campaigns-list-style shell command (it does not exist and will error, which is NOT a mureo bug).

No customer_id? (recovery)

Every Google Ads tool needs a customer_id (the target account). It is resolved from the stored credentials (set by mureo auth setup). When it is missing a tool returns customer_id is required. Provide it as a parameter or configure it … — auth is fine, only the account is unset (a common state when the operator finished sign-in but skipped account selection).

When you see that error, recover automatically — do NOT ask the operator to look up the ID in the Google Ads UI, and do NOT fall back to a CSV / "ask the agency":

1. Call google_ads_accounts_list — it needs no customer_id and returns every account reachable by the login.
2. Exactly one account → use its id as customer_id on the retried call and tell the operator which account you selected.
3. Several accounts → show the list and ask the operator which one to use.
4. Zero accounts → the login can't reach any Google Ads account; tell the operator to re-run mureo auth setup (or grant the account access), and that customer_id can be set there or via GOOGLE_ADS_CUSTOMER_ID (plus GOOGLE_ADS_LOGIN_CUSTOMER_ID for an MCC).

To persist the choice so it survives across sessions, point the operator at mureo auth setup (account picker) or the configure UI's env writer (GOOGLE_ADS_CUSTOMER_ID); passing customer_id per call works for the immediate request but is not remembered.

Tool Summary

API Resources

campaigns

• list -- List all campaigns in the account. Response includes channel_type ("SEARCH" | "DISPLAY" | ...) so you can tell search and display campaigns apart at a glance.

  Required: customer_id (string)
  Optional: status_filter (string: "ENABLED" | "PAUSED")

• get -- Get detailed information about a specific campaign. Response includes channel_type.

  Required: customer_id, campaign_id (string)

• create -- Create a search or display campaign. Requires user confirmation.

  Required: customer_id, name (string)
  Optional: bidding_strategy (string: "MAXIMIZE_CLICKS" | "MAXIMIZE_CONVERSIONS" | "TARGET_CPA" | ...),
            budget_id (string),
            channel_type (string: "SEARCH" | "DISPLAY"; defaults to "SEARCH")

  Note: For display campaigns, create the campaign with channel_type="DISPLAY", then create an ad group, then create the display ad via ads.create_display.
• update -- Update campaign settings (name, bidding strategy). Requires user confirmation.

  Required: customer_id, campaign_id (string)
  Optional: name (string), bidding_strategy (string)

• update_status -- Change campaign status. Requires user confirmation.

  Required: customer_id, campaign_id, status (string: "ENABLED" | "PAUSED" | "REMOVED")

• diagnose -- Comprehensive delivery diagnosis. Returns serving status, policy issues, budget constraints.

  Required: customer_id, campaign_id (string)

ad_groups

• list -- List ad groups, optionally filtered by campaign.

  Required: customer_id (string)
  Optional: campaign_id (string), status_filter (string)

• create -- Create an ad group within a campaign. Requires user confirmation.

  Required: customer_id, campaign_id, name (string)
  Optional: cpc_bid_micros (integer, in micros: 1,000,000 = 1 currency unit)

• update -- Update ad group settings. Requires user confirmation.

  Required: customer_id, ad_group_id (string)
  Optional: name (string), status (string: "ENABLED" | "PAUSED"), cpc_bid_micros (integer)

  Note: cpc_bid_micros is only accepted when the parent campaign uses a manual bidding strategy (MANUAL_CPC, MANUAL_CPM, MANUAL_CPV, ENHANCED_CPC). If the campaign uses an automated strategy (MAXIMIZE_CLICKS, MAXIMIZE_CONVERSIONS, TARGET_CPA, TARGET_ROAS, etc.), the tool returns a clear validation error — manual bids at the ad group level are not supported under automated bidding.

ads

• list -- List ads, optionally filtered by ad group. Returns headlines and descriptions for both RSA and RDA. For RDAs, the response also includes long_headline, business_name, marketing_images, square_marketing_images, and logo_images (lists of asset resource names). Performance Max campaigns return no rows here — P-MAX has no ad_group_ad. Its ad copy lives on asset groups: use asset_group_assets.list / asset_group_assets.replace below.

  Required: customer_id (string)
  Optional: ad_group_id (string), status_filter (string)

• create -- Create a Responsive Search Ad (RSA). Requires user confirmation.

  Required: customer_id, ad_group_id (string),
            headlines (array of strings, 3-15 items),
            descriptions (array of strings, 2-4 items)
  Optional: final_url (string), path1 (string), path2 (string)

  Note: For Responsive Display Ads, use ads.create_display instead.
• update -- Update RSA ad content. Requires user confirmation.

  Required: customer_id, ad_group_id, ad_id (string)
  Optional: headlines (array of strings), descriptions (array of strings)

  Note: This tool supports Responsive Search Ads (RSA) only. Calling it on a Responsive Display Ad (RDA) fails fast with a clear error — RDA text updates are not implemented; recreate the ad via ads.create_display instead.
• update_status -- Change ad status. Requires user confirmation.

  Required: customer_id, ad_group_id, ad_id, status (string: "ENABLED" | "PAUSED")

• policy_details -- Get ad policy approval details and disapproval reasons.

  Required: customer_id, ad_group_id, ad_id (string)

asset_group_assets (Performance Max ad copy)

A Performance Max campaign has no ad_group_ad, so ads.list / ads.update never see it. Its headlines, long headlines and descriptions are assets linked to an asset group. These two tools are the P-MAX ad-copy surface — do not tell an operator that swapping P-MAX copy is unsupported by the API; it is not.

• list -- List the HEADLINE / LONG_HEADLINE / DESCRIPTION assets linked to Performance Max asset groups. Returns one entry per link: {resource_name, field_type, status, asset_id, text, asset_group_id, asset_group_name, campaign_id, campaign_resource_name}, in the order the API returned them and not deduplicated. Start here when you have a campaign id but no asset group id.

  Required: customer_id (string)
  Optional: asset_group_id (string), campaign_id (string)

• replace -- Swap one headline / long headline / description for new text. Requires user confirmation.

  Required: customer_id, asset_group_id (string),
            field_type (string: "HEADLINE" | "LONG_HEADLINE" | "DESCRIPTION"),
            old_asset_id (string, from asset_group_assets.list),
            new_text (string)

  A Google Ads text Asset is immutable, so this is not an update: it creates a new Asset, links it under the same field_type, and removes the old link — all three in ONE atomic mutate, so the asset group's asset count for that field type never dips below the Performance Max minimum. Display-width limits (a full-width character counts as two): HEADLINE 30, LONG_HEADLINE 90, DESCRIPTION 90. Text already linked under the same field_type is refused before any write (Google rejects a duplicate link), as is an old_asset_id that is not linked under that field_type. The old Asset itself is not deleted, only its link to this asset group. Not automatically reversible: to undo, call replace again with the old text, which list reported.

keywords

• list -- List keywords with performance metrics.

  Required: customer_id (string)
  Optional: campaign_id (string), ad_group_id (string), status_filter (string)

• add -- Add keywords to an ad group. Requires user confirmation.

  Required: customer_id, ad_group_id (string),
            keywords (array of {text: string, match_type?: "BROAD" | "PHRASE" | "EXACT"})

  Note: match_type defaults to BROAD if omitted.
• remove -- Remove a keyword by criterion ID. Requires user confirmation.

  Required: customer_id, ad_group_id, criterion_id (string)

• suggest -- Get keyword suggestions from Keyword Planner.

  Required: customer_id (string), seed_keywords (array of strings)
  Optional: language_id (string, default: "1005" = Japanese),
            geo_id (string, default: "2392" = Japan)

• diagnose -- Diagnose keyword quality scores and delivery status for a campaign.

  Required: customer_id, campaign_id (string)

• pause -- Pause a keyword. Requires user confirmation.

  Required: customer_id, ad_group_id, criterion_id (string)

• audit -- Audit keyword performance and quality scores across a campaign.

  Required: customer_id, campaign_id (string)
  Optional: period (string, default: "LAST_30_DAYS"), target_cpa (number)

  period takes any date range constant or an explicit YYYY-MM-DD..YYYY-MM-DD / BETWEEN 'YYYY-MM-DD' AND 'YYYY-MM-DD' range.
• cross_adgroup_duplicates -- Find duplicate keywords across ad groups in a campaign.

  Required: customer_id, campaign_id (string)
  Optional: period (string, default: "LAST_30_DAYS")

  period takes any date range constant or an explicit YYYY-MM-DD..YYYY-MM-DD / BETWEEN 'YYYY-MM-DD' AND 'YYYY-MM-DD' range.

negative_keywords

• list -- List campaign-level negative keywords.

  Required: customer_id, campaign_id (string)

• add -- Add negative keywords to a campaign. Requires user confirmation.

  Required: customer_id, campaign_id (string),
            keywords (array of {text: string, match_type?: "BROAD" | "PHRASE" | "EXACT"})

• remove -- Remove a negative keyword. Requires user confirmation.

  Required: customer_id, campaign_id, criterion_id (string)

• add_to_ad_group -- Add negative keywords to an ad group. Requires user confirmation.

  Required: customer_id, ad_group_id (string),
            keywords (array of {text: string, match_type?: "BROAD" | "PHRASE" | "EXACT"})

• suggest -- Suggest negative keywords based on search term analysis.

  Required: customer_id, campaign_id (string)
  Optional: period (string, default: "LAST_30_DAYS"), target_cpa (number), ad_group_id (string)

  Period-over-period tool: fixed-length windows only (LAST_7_DAYS, LAST_14_DAYS, LAST_30_DAYS, LAST_90_DAYS, or an explicit range in either spelling).

negative_placements

Delivery-surface exclusions: excluded websites, mobile apps and mobile app categories. This is the placement side of exclusion; negative_keywords above is the search-term side. Both campaign-level and ad group-level criteria are supported — supply exactly one of campaign_id / ad_group_id.

A add call is recorded in STATE.json's action_log with an observation_due window and is reversible as one unit via rollback_apply, however many exclusions it carried. That is the whole point of routing placement hygiene through mureo: a bulk exclusion pass that kills delivery can be tied to a date and undone, instead of being reconstructed by hand.

• list -- List excluded websites / apps / app categories. Returns level ("campaign" | "ad_group"), criterion_id, type, criterion_type, value, display_name, and the parent campaign / ad group ids.

  Required: customer_id (string)
  Optional: campaign_id, ad_group_id (string)

• add -- Exclude websites / apps / app categories. Requires user confirmation. Always show the count and the level before writing — a large batch can take a Display campaign to zero impressions.

  Size it first (#547). Call analysis_exclusion_impact_preview with tool="google_ads_negative_placements_add" and the exact arguments you are about to send. It reports the share of the last N days' impressions / clicks / cost / conversions those placements carried, both for this batch and cumulatively for every standing exclusion once it lands, plus would_block. Show the operator the share, not just the count. If STRATEGY.md ## Guardrails carries max_delivery_share_removed_pct (or max_cumulative_delivery_share_removed_pct), an over-cap batch is refused before it reaches the API and the refusal names the measured share.

  Required: customer_id (string), exactly one of campaign_id / ad_group_id (string),
            placements (array of {type: "website" | "mobile_application" | "mobile_app_category",
                                  value: string})

  value formats: website = domain or URL; mobile_application = platform-prefixed app id (1- iOS, 2- Android); mobile_app_category = category constant id or its mobileAppCategoryConstants/<id> resource name.

• remove -- Lift exclusions by criterion_id. Requires user confirmation. Ids are verified first: anything that is not a negative placement criterion at the named level is reported under skipped and never removed.

  Required: customer_id (string), exactly one of campaign_id / ad_group_id (string),
            criterion_ids (array of string)

budget

• get -- Get the budget attached to a campaign, including its type. Returns id, name, daily_budget(_micros), total_budget / total_amount_micros (null unless a CUSTOM_PERIOD total budget), period (DAILY / CUSTOM_PERIOD), delivery_method, status, reference_count.

  Required: customer_id, campaign_id (string)

• update -- Update the daily and/or total budget amount. Requires user confirmation. Always show current vs. new amount. The period is immutable -- total amounts only apply to CUSTOM_PERIOD budgets.

  Required: customer_id, budget_id (string), and one of amount / amount_micros / total_amount / total_amount_micros

• create -- Create a new campaign budget. Requires user confirmation. Pass period=CUSTOM_PERIOD with total_amount(_micros) for a campaign-lifetime total budget (immutable after creation); otherwise supply the daily amount.

  Required: customer_id, name (string), and one of amount / total_amount / total_amount_micros
  Optional: period (DAILY | CUSTOM_PERIOD)

accounts

• list -- List all accessible Google Ads accounts under the manager account.

  Required: (none)

search_terms

• report -- Get search terms report showing actual queries that triggered ads.

  Required: customer_id (string)
  Optional: campaign_id (string), ad_group_id (string), period (string, default: "LAST_30_DAYS")

• review -- Multi-rule review of search terms with add/exclude recommendations.

  Required: customer_id, campaign_id (string)
  Optional: period (string, default: "LAST_7_DAYS"), target_cpa (number)

• analyze -- Analyze search terms with intent classification and clustering.

  Required: customer_id, campaign_id (string)

sitelinks

• list -- List sitelink extensions for a campaign.

  Required: customer_id, campaign_id (string)

• create -- Create a sitelink extension. Requires user confirmation.

  Required: customer_id, campaign_id, link_text, final_url (string)

• remove -- Remove a sitelink extension. Requires user confirmation.

  Required: customer_id, campaign_id, extension_id (string)

callouts

• list -- List callout extensions for a campaign.

  Required: customer_id, campaign_id (string)

• create -- Create a callout extension. Requires user confirmation.

  Required: customer_id, campaign_id, callout_text (string)

• remove -- Remove a callout extension. Requires user confirmation.

  Required: customer_id, campaign_id, extension_id (string)

conversions

• list -- List all conversion actions in the account.

  Required: customer_id (string)

• get -- Get details of a specific conversion action.

  Required: customer_id, conversion_action_id (string)

• performance -- Get conversion performance metrics across campaigns.

  Required: customer_id (string)

• create -- Create a new conversion action. Requires user confirmation.

  Required: customer_id, name, type (string)

• update -- Update a conversion action. Requires user confirmation.

  Required: customer_id, conversion_action_id (string)

• remove -- Remove a conversion action. Requires user confirmation.

  Required: customer_id, conversion_action_id (string)

• tag -- Get the conversion tracking tag (JavaScript snippet) for a conversion action.

  Required: customer_id, conversion_action_id (string)

targeting

• recommendations.list -- List optimization recommendations from Google.

  Required: customer_id (string)

• recommendations.apply -- Apply an optimization recommendation. Requires user confirmation.

  Required: customer_id, recommendation_id (string)

• device_targeting.get -- Get device targeting bid adjustment settings.

  Required: customer_id, campaign_id (string)

• device_targeting.set -- Set device targeting bid adjustments. Requires user confirmation.

  Required: customer_id, campaign_id, device_type, bid_modifier (string/number)

• bid_adjustments.get -- Get bid adjustment settings for a campaign.

  Required: customer_id, campaign_id (string)

• bid_adjustments.update -- Update bid adjustments. Requires user confirmation.

  Required: customer_id, campaign_id (string)

• location_targeting.list -- List location targeting criteria.

  Required: customer_id, campaign_id (string)

• location_targeting.update -- Update location targeting settings. Requires user confirmation.

  Required: customer_id, campaign_id (string)

• schedule_targeting.list -- List ad schedule targeting rules.

  Required: customer_id, campaign_id (string)

• schedule_targeting.update -- Update ad schedule targeting. Requires user confirmation.

  Required: customer_id, campaign_id (string)

• change_history.list -- List account change history (who changed what, when).

  Required: customer_id (string)

• demographic_targeting.list -- List explicit demographic criteria (age range, gender, parental status, income). Excluded segments have negative=true; segments with no explicit criterion are targeted by default and not returned.

  Required: customer_id (string)
  Optional: ad_group_id (string), campaign_id (string)

• audience_targeting.list -- List audience-type criteria (user interests, user lists, custom/combined audiences). value is the criterion's resource name.

  Required: customer_id (string)
  Optional: ad_group_id (string), campaign_id (string)

analysis & reporting

• performance.report -- Get performance metrics aggregated by campaign.

  Required: customer_id (string)
  Optional: campaign_id (string), period (string, default: "LAST_30_DAYS")

  Period options: TODAY, YESTERDAY, THIS_WEEK_SUN_TODAY, THIS_WEEK_MON_TODAY, LAST_BUSINESS_WEEK, LAST_WEEK_SUN_SAT, LAST_WEEK_MON_SUN, LAST_7_DAYS, LAST_14_DAYS, LAST_30_DAYS, LAST_90_DAYS, THIS_MONTH, LAST_MONTH — or an explicit range: YYYY-MM-DD..YYYY-MM-DD or BETWEEN 'YYYY-MM-DD' AND 'YYYY-MM-DD'
• performance.analyze -- Analyze performance trends, detect anomalies, and surface insights.

  Required: customer_id (string)

• cost_increase.investigate -- Investigate sudden cost increases and identify root causes.

  Required: customer_id, campaign_id (string)

• health_check.all -- Run a comprehensive account health check across all campaigns.

  Required: customer_id (string)

• ad_performance.compare -- Compare ad performance across variants in an ad group.

  Required: customer_id, ad_group_id (string)

• ad_performance.report -- Get detailed ad-level performance report.

  Required: customer_id (string)

• network_performance.report -- Get performance breakdown by network (Search, Display, etc.).

  Required: customer_id (string)

• budget.efficiency -- Analyze budget utilization efficiency across campaigns.

  Required: customer_id (string)

• budget.reallocation -- Suggest budget reallocation across campaigns based on performance.

  Required: customer_id (string)

• auction_insights.get -- Get auction insights with competitor impression share and overlap rate.

  Required: customer_id, campaign_id (string)

• auction_insights.analyze -- Analyze auction insights: impression share, overlap rate, top-of-page rate, and competitor domains.

  Required: customer_id, campaign_id (string)
  Optional: period (string, default: "LAST_30_DAYS")

• rsa_assets.analyze -- Analyze RSA asset (headline/description) performance ratings.

  Required: customer_id, ad_group_id (string)

• rsa_assets.audit -- Audit RSA assets for best practices and coverage gaps.

  Required: customer_id, campaign_id (string)

• cpc.detect_trend -- Detect CPC trend direction (rising / stable / falling) using linear regression on daily CPC data. Also detects sudden spikes and week-over-week changes.

  Required: customer_id, campaign_id (string)
  Optional: period (string, default: "LAST_30_DAYS")

• device.analyze -- Compare performance across devices (Desktop / Mobile / Tablet). Shows CPA, CVR, CTR per device, detects CPA gaps, and identifies wasted spend on zero-conversion devices.

  Required: customer_id, campaign_id (string)
  Optional: period (string, default: "LAST_30_DAYS")

b2b

• btob.optimizations -- Get B2B-specific optimization suggestions (lead quality, audience targeting, etc.).

  Required: customer_id (string)

creative

• landing_page.analyze -- Analyze landing page relevance and quality signals for a campaign.

  Required: customer_id, campaign_id (string)

• creative.research -- Research competitive creative strategies and ad copy patterns.

  Required: customer_id (string)

monitoring

• monitoring.delivery_goal -- Monitor campaign delivery pace against goals.

  Required: customer_id, campaign_id (string)

• monitoring.cpa_goal -- Monitor CPA against target goals and alert on deviations.

  Required: customer_id, campaign_id (string)

• monitoring.cv_goal -- Monitor conversion volume against goals.

  Required: customer_id, campaign_id (string)

• monitoring.zero_conversions -- Detect campaigns with zero conversions in recent periods.

  Required: customer_id (string)

capture

• capture.screenshot -- Capture a screenshot of a URL (useful for landing page audits).

  Required: url (string)

assets

• assets.upload_image -- Upload a local image file as a Google Ads asset.

  Required: customer_id, file_path (string)

• image_assets.list -- List image assets with names, dimensions, and serving URLs.

  Required: customer_id (string)
  Optional: limit (integer 1-1000, default: 100)

Common Workflows

1. Performance Check

A typical daily performance check flow:

Step 1: List campaigns to identify active ones
  -> google_ads_campaigns_list {customer_id, status_filter: "ENABLED"}

Step 2: Get overall performance report
  -> google_ads_performance_report {customer_id, period: "LAST_7_DAYS"}

Step 3: Run comprehensive health check
  -> google_ads_health_check_all {customer_id}

Step 4: For underperforming campaigns, run diagnostics
  -> google_ads_campaigns_diagnose {customer_id, campaign_id}

Step 5: Check keyword quality scores
  -> google_ads_keywords_diagnose {customer_id, campaign_id}

Step 6: Analyze device performance for CPA gaps
  -> google_ads_device_analyze {customer_id, campaign_id}

Step 7: Monitor conversion goals
  -> google_ads_monitoring_zero_conversions {customer_id}

2. Negative Keyword Addition

When cleaning up wasted spend from irrelevant search terms:

Step 1: Get search terms report
  -> google_ads_search_terms_report {customer_id, campaign_id, period: "LAST_30_DAYS"}

Step 2: Or use automated review for suggestions
  -> google_ads_search_terms_review {customer_id, campaign_id, target_cpa: 3000}

Step 3: Analyze search terms with intent classification
  -> google_ads_search_terms_analyze {customer_id, campaign_id}

Step 4: Get AI-suggested negative keywords
  -> google_ads_negative_keywords_suggest {customer_id, campaign_id}

Step 5: Review current negative keywords to avoid duplicates
  -> google_ads_negative_keywords_list {customer_id, campaign_id}

Step 6: Add negative keywords (CONFIRM WITH USER)
  -> google_ads_negative_keywords_add {customer_id, campaign_id, keywords: [...]}

3. New Campaign Creation

Full campaign setup from scratch:

Step 1: Create a budget
  -> google_ads_budget_create {customer_id, name, amount}

Step 2: Create the campaign (omit channel_type or set "SEARCH" for a search campaign)
  -> google_ads_campaigns_create {customer_id, name, bidding_strategy, budget_id}

Step 3: Create ad groups
  -> google_ads_ad_groups_create {customer_id, campaign_id, name}

Step 4: Add keywords to each ad group
  -> google_ads_keywords_add {customer_id, ad_group_id, keywords: [...]}

Step 5: Create RSA ads for each ad group
  -> analysis_tracking_consistency_check {ads: <existing ads in the campaign>, planned_ads: [...]}
     (tracking-parameter pre-flight -- see ../_mureo-shared/SKILL.md; stop on any finding)
  -> google_ads_ads_create {customer_id, ad_group_id, headlines: [...], descriptions: [...], final_url}

Step 6: Add campaign-level negative keywords
  -> google_ads_negative_keywords_add {customer_id, campaign_id, keywords: [...]}

Step 7: Add sitelink extensions
  -> google_ads_sitelinks_create {customer_id, campaign_id, link_text, final_url}

Step 8: Add callout extensions
  -> google_ads_callouts_create {customer_id, campaign_id, callout_text}

Step 9: Set up conversion tracking
  -> google_ads_conversions_create {customer_id, name, type}

Step 10: Enable the campaign
  -> google_ads_campaigns_update_status {customer_id, campaign_id, status: "ENABLED"}

3b. New Display Campaign + Responsive Display Ad

Display campaigns use a different channel_type and require image assets. mureo uploads the local image files automatically before creating the ad.

Step 1: Create a budget
  -> google_ads_budget_create {customer_id, name, amount}

Step 2: Create a DISPLAY campaign
  -> google_ads_campaigns_create {
       customer_id, name, channel_type: "DISPLAY",
       bidding_strategy: "MAXIMIZE_CONVERSIONS", budget_id
     }

Step 3: Create an ad group inside the display campaign
  -> google_ads_ad_groups_create {customer_id, campaign_id, name}

Step 4: Create the responsive display ad
  -> google_ads_ads_create_display {
       customer_id, ad_group_id,
       headlines: [...], long_headline, descriptions: [...],
       business_name,
       marketing_image_paths: ["/path/to/marketing-1200x628.jpg"],
       square_marketing_image_paths: ["/path/to/square-1200x1200.jpg"],
       logo_image_paths: ["/path/to/logo.png"],   # optional
       final_url
     }

Step 5: Enable the campaign
  -> google_ads_campaigns_update_status {customer_id, campaign_id, status: "ENABLED"}

Constraints (RDA):

• Headlines: 1-5, each ≤30 display width
• Long headline: required, ≤90 display width
• Descriptions: 1-5, each ≤90 display width
• Business name: required, ≤25 display width
• Marketing images (1.91:1): 1-15, 3+ recommended for delivery quality
• Square marketing images (1:1): 1-15, 3+ recommended
• Logo images: optional, up to 5
• The target ad group must belong to a DISPLAY campaign — mureo verifies this before any upload happens to avoid orphaned assets.
• If image upload or ad creation fails partway through, an RDAUploadError is raised that includes the resource names of any orphaned uploaded assets so the agent can clean them up.

4. Budget Adjustment

Safe budget change workflow:

Step 1: Get current budget
  -> google_ads_budget_get {customer_id, campaign_id}

Step 2: Analyze budget efficiency
  -> google_ads_budget_efficiency {customer_id}

Step 3: Get reallocation suggestions
  -> google_ads_budget_reallocation {customer_id}

Step 4: Update budget (CONFIRM WITH USER)
  -> google_ads_budget_update {customer_id, budget_id, amount: 6000}

5. Delivery Troubleshooting

When a campaign is not getting impressions:

Step 1: Check campaign status and settings
  -> google_ads_campaigns_get {customer_id, campaign_id}

Step 2: Run comprehensive delivery diagnosis
  -> google_ads_campaigns_diagnose {customer_id, campaign_id}

Step 3: Check keyword quality and delivery status
  -> google_ads_keywords_diagnose {customer_id, campaign_id}

Step 4: Check ad policy details for disapprovals
  -> google_ads_ads_policy_details {customer_id, ad_group_id, ad_id}

Step 5: Check CPC trends (are bids competitive?)
  -> google_ads_cpc_detect_trend {customer_id, campaign_id}

Step 6: Analyze auction insights (competitor landscape)
  -> google_ads_auction_insights_analyze {customer_id, campaign_id}

Step 7: Monitor delivery pace
  -> google_ads_monitoring_delivery_goal {customer_id, campaign_id}

Step 8: Review budget constraints
  -> google_ads_budget_get {customer_id, campaign_id}

6. Competitive Analysis

Understanding the competitive landscape:

Step 1: Analyze auction insights
  -> google_ads_auction_insights_analyze {customer_id, campaign_id, period: "LAST_30_DAYS"}

Step 2: Get auction insights data
  -> google_ads_auction_insights_get {customer_id, campaign_id}

Step 3: Detect CPC trend (rising CPC may indicate increased competition)
  -> google_ads_cpc_detect_trend {customer_id, campaign_id}

Step 4: Investigate cost increases
  -> google_ads_cost_increase_investigate {customer_id, campaign_id}

Step 5: Research competitive creative strategies
  -> google_ads_creative_research {customer_id}

Step 6: Get keyword suggestions to find new opportunities
  -> google_ads_keywords_suggest {customer_id, seed_keywords: ["..."]}

7. RSA Optimization

Improving responsive search ad performance:

Step 1: Analyze RSA asset performance
  -> google_ads_rsa_assets_analyze {customer_id, campaign_id}

Step 2: Audit RSA assets for best practices
  -> google_ads_rsa_assets_audit {customer_id, campaign_id}

Step 3: Compare ad variants
  -> google_ads_ad_performance_compare {customer_id, ad_group_id}

Step 4: Check landing page relevance
  -> google_ads_landing_page_analyze {customer_id, url}

8. Account Audit

Comprehensive account review:

Step 1: List all accessible accounts
  -> google_ads_accounts_list {}

Step 2: Run health check
  -> google_ads_health_check_all {customer_id}

Step 3: Audit keywords
  -> google_ads_keywords_audit {customer_id, campaign_id}

Step 4: Find duplicate keywords
  -> google_ads_keywords_cross_adgroup_duplicates {customer_id, campaign_id}

Step 5: Review change history
  -> google_ads_change_history_list {customer_id}

Step 6: Check optimization recommendations
  -> google_ads_recommendations_list {customer_id}

Important Notes

• Micros: Some values (e.g., cpc_bid_micros) are in micros where 1,000,000 = 1 currency unit. For example, 500,000 micros = 0.50 in the account currency.
• Period values: Use Google Ads date range constants like LAST_7_DAYS, LAST_30_DAYS, LAST_90_DAYS, THIS_MONTH. For a window no constant reaches — a single past calendar month, say — pass an explicit range instead: period: "2026-05-01..2026-05-31" — the YYYY-MM-DD..YYYY-MM-DD form Meta Ads takes too — or, in GAQL spelling, period: "BETWEEN '2026-05-01' AND '2026-05-31'" (both endpoints inclusive, account time zone).
• LAST_90_DAYS uses the server date: every other constant is resolved by Google Ads in the account's reporting time zone, but LAST_90_DAYS has no API constant — mureo expands it into the 90 days ending yesterday on the server's date. When the server and the account sit in different time zones the window edges can differ by a day. Pass an explicit range when the exact boundary matters.
• Period-over-period tools (search_terms.review, performance.analyze, negative_keywords.suggest) also read the equal-length window immediately before the one you ask for, so they take only fixed-length windows: LAST_7_DAYS, LAST_14_DAYS, LAST_30_DAYS, LAST_90_DAYS, or an explicit range in either spelling. Calendar constants such as THIS_MONTH are refused rather than quietly replaced with a different window.
• customer_id: Always a 10-digit string without dashes (e.g., "1234567890", not "123-456-7890").
• RSA limits: Headlines 3-15, descriptions 2-4. Maximum 3 enabled RSA ads per ad group.
• Write operations: All tools that create, update, or remove resources require user confirmation before execution.