All 160 MCP tools. Metadata Developer Docs

archive_audience Destructive Archive (delete) a custom audience from the Metadata platform.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`custom_audience_id`	integer		The numeric ID of the custom audience to archive.
`audience_name`	string		The name of the custom audience to archive. If provided without custom_audience_id, the tool will search for the audience by name to resolve the ID.

Full page, schema and recorded response · machine index

create_audience_from_segment Writes Create a new custom audience based on a segment from the insights/attribution tool.

Creates or changes something in the account.

Parameter	Type		Notes
`segment_id`	integer	required	The ID of the segment from list_segments tool. Use list_segments to find available segment IDs.
`name`	string	required	Name for the new audience. Must be unique and descriptive. Example: 'High Intent Accounts - 30 Days'
`lookback_window`	integer	required	Number of days to look back in the segment results. Range: 1 to 365 days. Example: 30 for last 30 days, 90 for last quarter, 365 for last year.

Full page, schema and recorded response · machine index

create_bombora_audience Writes Create Bombora buyer intent audiences on the Metadata platform.

Creates or changes something in the account.

Parameter	Type		Notes
`name`	string	required	Audience name (required)
`topics`	array	required	Array of buyer intent topics (required). Use get_intent_topics to retrieve available topics. Example: ['Pardot', 'Salesforce']
`industries`	array		Industry IDs for company targeting (optional). Valid range: 1-148. Example: [4, 96]
`employees`	array		Employee count ranges for company targeting (optional). Valid values: 'FROM_1_TO_10', 'FROM_10_TO_50', 'FROM_50_TO_200', 'FROM_200_TO_500', 'FROM_500_TO_1000', 'FROM_1000_TO_5000', 'FROM_5000_TO_10000', 'OVER_10000'
`location_country_ids`	array		Country IDs for geographic targeting (optional). Valid IDs: 1 (Afghanistan), 2 (Albania), 3 (Algeria), 4 (American Samoa), 5 (Andorra), 6 (Angola), 7 (Anguilla), 8 (Antarctica), 9 (Antigua and Barbuda), 10 (Argentina), 11 (Armenia), 12 (Aruba), 13 (Australia), 14 (Austria), 15 (Azerbaijan), 16 (Baha
`location_state_ids`	array		State IDs for geographic targeting (optional). Valid US State IDs: 1 (Alabama), 2 (Alaska), 3 (Arizona), 4 (Arkansas), 5 (California), 6 (Colorado), 7 (Connecticut), 8 (Delaware), 9 (Florida), 10 (Georgia), 11 (Hawaii), 12 (Idaho), 13 (Illinois), 14 (Indiana), 15 (Iowa), 16 (Kansas), 17 (Kentucky),
`job_title_includes`	array		Free text keywords for job titles to include (optional). Example: ['Software Engineer', 'Manager']
`job_title_excludes`	array		Free text keywords for job titles to exclude (optional). Example: ['Senior', 'Lead']
`job_function_include_ids`	array		Job function IDs to include (optional). Valid IDs: 1 (Marketing), 2 (Sales), 3 (Customer Service), 4 (Finance), 5 (HR), 6 (PR), 7 (Media), 8 (Design), 9 (Engineering), 10 (Operations), 11 (Legal), 12 (Education), 13 (Health), 14 (Real Estate), 15 (Accounting), 16 (Administrative), 17 (Business Devel
`job_function_exclude_ids`	array		Job function IDs to exclude (optional). Valid IDs: 1-28 as listed above
`seniority_include_ids`	array		Seniority level IDs to include (optional). Valid IDs: 2 (Manager), 3 (Director), 4 (VP), 5 (CXO), 6 (Unpaid), 7 (Training), 8 (Senior), 9 (Owner), 10 (Entry), 11 (Partner). Example: [10, 7] for Entry and Training
`seniority_exclude_ids`	array		Seniority level IDs to exclude (optional). Valid IDs: 2-11 as listed above
`contacts_per_company_limit`	integer		Maximum number of contacts to target per company (optional). Example: 20

Full page, schema and recorded response · machine index

create_csv_upload_linkedin_native_audience Writes Create a **CSV Upload - LinkedIn Native** audience (platform `customAudienceType=NATIVE_TARGETING_CSV`).

Creates or changes something in the account.

Parameter	Type		Notes
`audience_name`	string	required	Name for the new audience (required).
`companies`	object		Inline map of company names to website URLs (optional). MUTUALLY EXCLUSIVE with `companies_source_csv_url`. Example: {"Acme Corp": "https://acme.com"}
`companies_source_csv_url`	string		Public URL of a CSV with header EXACTLY `companyname,companywebsite` (case-insensitive). Use when the user attached a CSV to the chat — the URL comes via `AudienceBrief.attached_file_urls`. MUTUALLY EXCLUSIVE with `companies`.
`location_country_ids`	array		Country IDs for geographic targeting (optional). Valid IDs: 1 (Afghanistan), 2 (Albania), 3 (Algeria), 4 (American Samoa), 5 (Andorra), 6 (Angola), 7 (Anguilla), 8 (Antarctica), 9 (Antigua and Barbuda), 10 (Argentina), 11 (Armenia), 12 (Aruba), 13 (Australia), 14 (Austria), 15 (Azerbaijan), 16 (Baha
`employees`	array		LinkedIn employee count ranges for audience targeting. Valid values: 1, 2-10, 11-50, 51-200, 201-500, 501-1000, 1001-5000, 5001-10000, +10001. Example: ['201-500', '501-1000']
`revenues`	array		LinkedIn revenue ranges for audience targeting. Valid values: Under $1M, $1M-$10M, $10M-$100M, $100M-$1B, $1B+. Example: ['$1M-$10M', '$10M-$100M']
`company_names`	array		Free-text company names (resolved to LinkedIn company IDs). Example: ['Metadata', 'Google']
`job_titles`	array		Free-text job titles (resolved to LinkedIn job-title IDs). Example: ['Software Engineer', 'Product Manager']
`skills`	array		Free-text professional skills (resolved to LinkedIn skill IDs). Example: ['HubSpot', 'Salesforce']

Full page, schema and recorded response · machine index

create_facebook_native_criteria_audience Writes Create a Facebook native criteria audience using Facebook-supported targeting criteria.

Creates or changes something in the account.

Parameter	Type		Notes
`name`	string	required	Audience name (required).
`industries`	array		Facebook native industry categories for audience targeting. Valid values: Administrative Services, Architecture and Engineering, Arts, Entertainment, Sports and Media, Business and Finance, Business decision maker titles and interests, Cleaning and Maintenance Services, Community and Social Services
`countries`	array		List of country names to target. Each must be a valid country name (e.g., ['United States', 'Brazil']). Resolved to platform country IDs before submission.
`job_titles`	array		Free-text job title strings for targeting. Example: ['Software Engineer', 'Senior Developer', 'Product Owner']
`interests`	array		Free-text interest strings for targeting. Example: ['Java', 'Python', 'Cloud Computing', 'Machine Learning']
`min_age`	integer	required	Minimum age for audience targeting. Must be between 21 and 65, and less than or equal to max_age.
`max_age`	integer	required	Maximum age for audience targeting. Must be between 21 and 65, and greater than or equal to min_age.
`estimate_only`	boolean		If true, do NOT create the audience — only estimate its size and return the expected contact count. Use this when the user wants to preview Facebook audience size before committing. Defaults to false (audience is created).

Full page, schema and recorded response · machine index

create_firmographic_audience Writes Create B2B company firmographic audiences for targeted advertising campaigns.

Creates or changes something in the account.

Parameter	Type		Notes
`name`	string	required	Audience name (required)
`industries`	array		Industry IDs for firmographic/technographic filtering (optional). Valid IDs: 1=Defense & Space, 3=Computer Hardware, 4=Computer Software, 5=Computer Networking, 6=Internet, 7=Semiconductors, 8=Telecommunications, 9=Law Practice, 10=Legal Services, 11=Management Consulting, 12=Biotechnology, 13=Medic
`employees`	array		Employee count ranges. Valid enum values: LESS_THAN_10 (0-9), FROM_10_TO_50 (10-49), FROM_50_TO_200 (50-199), FROM_200_TO_500 (200-499), FROM_500_TO_1000 (500-999), FROM_1000_TO_5000 (1,000-4,999), FROM_5000_TO_10000 (5,000-9,999), OVER_10000 (10,000+)
`revenue`	array		Revenue ranges. Valid enum values: LESS_THAN_1M ($0-$1M), FROM_1M_TO_10M ($1M-$10M), FROM_10M_TO_50M ($10M-$50M), FROM_50M_TO_100M ($50M-$100M), FROM_100M_TO_200M ($100M-$200M), FROM_200M_TO_500M ($200M-$500M), FROM_500M_TO_1000M ($500M-$1B), OVER_1000M ($1B+)
`location_country_ids`	array		Country IDs for geographic targeting (optional). Valid IDs: 1 (Afghanistan), 2 (Albania), 3 (Algeria), 4 (American Samoa), 5 (Andorra), 6 (Angola), 7 (Anguilla), 8 (Antarctica), 9 (Antigua and Barbuda), 10 (Argentina), 11 (Armenia), 12 (Aruba), 13 (Australia), 14 (Austria), 15 (Azerbaijan), 16 (Baha
`location_state_ids`	array		State IDs for geographic targeting (optional). Valid US State IDs: 1 (Alabama), 2 (Alaska), 3 (Arizona), 4 (Arkansas), 5 (California), 6 (Colorado), 7 (Connecticut), 8 (Delaware), 9 (Florida), 10 (Georgia), 11 (Hawaii), 12 (Idaho), 13 (Illinois), 14 (Indiana), 15 (Iowa), 16 (Kansas), 17 (Kentucky),
`job_title_includes`	array		Free text keywords for job titles to include (optional). Example: ['Software Engineer', 'Manager']
`job_title_excludes`	array		Free text keywords for job titles to exclude (optional). Example: ['Senior', 'Lead']
`job_function_include_ids`	array		Job function IDs to include (optional). Valid IDs: 1 (Marketing), 2 (Sales), 3 (Customer Service), 4 (Finance), 5 (HR), 6 (PR), 7 (Media), 8 (Design), 9 (Engineering), 10 (Operations), 11 (Legal), 12 (Education), 13 (Health), 14 (Real Estate), 15 (Accounting), 16 (Administrative), 17 (Business Devel
`job_function_exclude_ids`	array		Job function IDs to exclude (optional). Valid IDs: 1-28 as listed above
`seniority_include_ids`	array		Seniority level IDs to include (optional). Valid IDs: 2 (Manager), 3 (Director), 4 (VP), 5 (CXO), 6 (Unpaid), 7 (Training), 8 (Senior), 9 (Owner), 10 (Entry), 11 (Partner). Example: [10, 7] for Entry and Training
`seniority_exclude_ids`	array		Seniority level IDs to exclude (optional). Valid IDs: 2-11 as listed above
`contacts_per_company_limit`	integer		Maximum number of contacts to target per company (optional). Example: 20
`estimate_only`	boolean		If true, do NOT create the audience — only estimate its size and return the counts. Use this when the user wants to preview how many companies/contacts match the criteria before committing. Defaults to false (audience is created).

Full page, schema and recorded response · machine index

create_g2_intent_dynamic_audience Writes Create a **G2 Intent (Dynamic)** audience (platform `customAudienceType=DYNAMIC_G2_CONTACT`).

Creates or changes something in the account.

Parameter	Type		Notes
`name`	string	required	Audience name (required, max 50 characters).
`intent_days`	integer	required	Days to look back for G2 intent signals (required, 1-365).
`buying_stages`	array		Buying stages to target. The platform REQUIRES at least one; if the user did not specify, ASK them before calling — do NOT default.
`location_country_ids`	array		Country IDs for geographic targeting (optional). Valid IDs: 1 (Afghanistan), 2 (Albania), 3 (Algeria), 4 (American Samoa), 5 (Andorra), 6 (Angola), 7 (Anguilla), 8 (Antarctica), 9 (Antigua and Barbuda), 10 (Argentina), 11 (Armenia), 12 (Aruba), 13 (Australia), 14 (Austria), 15 (Azerbaijan), 16 (Baha
`location_state_ids`	array		State IDs for geographic targeting (optional). Valid US State IDs: 1 (Alabama), 2 (Alaska), 3 (Arizona), 4 (Arkansas), 5 (California), 6 (Colorado), 7 (Connecticut), 8 (Delaware), 9 (Florida), 10 (Georgia), 11 (Hawaii), 12 (Idaho), 13 (Illinois), 14 (Indiana), 15 (Iowa), 16 (Kansas), 17 (Kentucky),
`job_title_includes`	array		Free text keywords for job titles to include (optional). Example: ['Software Engineer', 'Manager']
`job_title_excludes`	array		Free text keywords for job titles to exclude (optional). Example: ['Senior', 'Lead']
`job_function_include_ids`	array		Job function IDs to include (optional). Valid IDs: 1 (Marketing), 2 (Sales), 3 (Customer Service), 4 (Finance), 5 (HR), 6 (PR), 7 (Media), 8 (Design), 9 (Engineering), 10 (Operations), 11 (Legal), 12 (Education), 13 (Health), 14 (Real Estate), 15 (Accounting), 16 (Administrative), 17 (Business Devel
`job_function_exclude_ids`	array		Job function IDs to exclude (optional). Valid IDs: 1-28 as listed above
`seniority_include_ids`	array		Seniority level IDs to include (optional). Valid IDs: 2 (Manager), 3 (Director), 4 (VP), 5 (CXO), 6 (Unpaid), 7 (Training), 8 (Senior), 9 (Owner), 10 (Entry), 11 (Partner). Example: [10, 7] for Entry and Training
`seniority_exclude_ids`	array		Seniority level IDs to exclude (optional). Valid IDs: 2-11 as listed above
`contacts_per_company_limit`	integer		Maximum number of contacts to target per company (optional). Example: 20

Full page, schema and recorded response · machine index

create_g2_intent_linkedin_native_dynamic_audience Writes Create a **G2 Intent - LinkedIn Native (Dynamic)** audience (platform `customAudienceType=DYNAMIC_G2`).

Creates or changes something in the account.

Parameter	Type		Notes
`name`	string	required	Audience name (required, max 50 characters).
`intent_days`	integer	required	Days to look back for G2 intent signals (required, 1-365).
`buying_stages`	array		Buying stages to target. The platform REQUIRES at least one; if the user did not specify, ASK them before calling — do NOT default.
`employees`	array		LinkedIn employee count ranges for audience targeting. Valid values: 1, 2-10, 11-50, 51-200, 201-500, 501-1000, 1001-5000, 5001-10000, +10001. Example: ['201-500', '501-1000']
`revenues`	array		LinkedIn revenue ranges for audience targeting. Valid values: Under $1M, $1M-$10M, $10M-$100M, $100M-$1B, $1B+. Example: ['$1M-$10M', '$10M-$100M']
`company_names`	array		Free-text company names to target (resolved server-side to LinkedIn company IDs). Example: ['Metadata', 'Google', 'Salesforce']
`location_country_ids`	array		Country IDs for geographic targeting (optional). Valid IDs: 1 (Afghanistan), 2 (Albania), 3 (Algeria), 4 (American Samoa), 5 (Andorra), 6 (Angola), 7 (Anguilla), 8 (Antarctica), 9 (Antigua and Barbuda), 10 (Argentina), 11 (Armenia), 12 (Aruba), 13 (Australia), 14 (Austria), 15 (Azerbaijan), 16 (Baha
`job_titles`	array		Free-text job titles (resolved to LinkedIn job-title IDs). Example: ['Software Engineer', 'Product Manager']
`skills`	array		Free-text professional skills (resolved to LinkedIn skill IDs). Example: ['HubSpot', 'Salesforce', 'Python']

Full page, schema and recorded response · machine index

create_g2_intent_static_audience Writes Create a **G2 Intent (Static)** audience (platform `customAudienceType=G2_ABM_CONTACT`).

Creates or changes something in the account.

Parameter	Type		Notes
`name`	string	required	Audience name (required, max 50 characters).
`intent_days`	integer	required	Days to look back for G2 intent signals (required, 1-365).
`buying_stages`	array		Buying stages to target. The platform REQUIRES at least one; if the user did not specify, ASK them before calling — do NOT default.
`location_country_ids`	array		Country IDs for geographic targeting (optional). Valid IDs: 1 (Afghanistan), 2 (Albania), 3 (Algeria), 4 (American Samoa), 5 (Andorra), 6 (Angola), 7 (Anguilla), 8 (Antarctica), 9 (Antigua and Barbuda), 10 (Argentina), 11 (Armenia), 12 (Aruba), 13 (Australia), 14 (Austria), 15 (Azerbaijan), 16 (Baha
`location_state_ids`	array		State IDs for geographic targeting (optional). Valid US State IDs: 1 (Alabama), 2 (Alaska), 3 (Arizona), 4 (Arkansas), 5 (California), 6 (Colorado), 7 (Connecticut), 8 (Delaware), 9 (Florida), 10 (Georgia), 11 (Hawaii), 12 (Idaho), 13 (Illinois), 14 (Indiana), 15 (Iowa), 16 (Kansas), 17 (Kentucky),
`job_title_includes`	array		Free text keywords for job titles to include (optional). Example: ['Software Engineer', 'Manager']
`job_title_excludes`	array		Free text keywords for job titles to exclude (optional). Example: ['Senior', 'Lead']
`job_function_include_ids`	array		Job function IDs to include (optional). Valid IDs: 1 (Marketing), 2 (Sales), 3 (Customer Service), 4 (Finance), 5 (HR), 6 (PR), 7 (Media), 8 (Design), 9 (Engineering), 10 (Operations), 11 (Legal), 12 (Education), 13 (Health), 14 (Real Estate), 15 (Accounting), 16 (Administrative), 17 (Business Devel
`job_function_exclude_ids`	array		Job function IDs to exclude (optional). Valid IDs: 1-28 as listed above
`seniority_include_ids`	array		Seniority level IDs to include (optional). Valid IDs: 2 (Manager), 3 (Director), 4 (VP), 5 (CXO), 6 (Unpaid), 7 (Training), 8 (Senior), 9 (Owner), 10 (Entry), 11 (Partner). Example: [10, 7] for Entry and Training
`seniority_exclude_ids`	array		Seniority level IDs to exclude (optional). Valid IDs: 2-11 as listed above
`contacts_per_company_limit`	integer		Maximum number of contacts to target per company (optional). Example: 20

Full page, schema and recorded response · machine index

create_linkedin_native_criteria_audience Writes Create a LinkedIn native criteria audience using LinkedIn-supported targeting criteria.

Creates or changes something in the account.

Parameter	Type		Notes
`name`	string	required	Audience name (required, must be shorter than 50 characters).
`job_titles`	array		Free-text job title strings for targeting. Example: ['Software Engineer', 'Product Manager']
`skills`	array		Free-text skill strings for targeting. Example: ['HubSpot', 'Salesforce', 'Python']
`company_names`	array		Free-text company name strings for targeting. Example: ['Metadata', 'Google', 'Salesforce']
`employees`	array		LinkedIn employee count ranges for audience targeting. Valid values: 1, 2-10, 11-50, 51-200, 201-500, 501-1000, 1001-5000, 5001-10000, +10001. Example: ['201-500', '501-1000']
`revenues`	array		LinkedIn revenue ranges for audience targeting. Valid values: Under $1M, $1M-$10M, $10M-$100M, $100M-$1B, $1B+. Example: ['$1M-$10M', '$10M-$100M']
`location_country_ids`	array		Country IDs for geographic targeting (optional). Valid IDs: 1 (Afghanistan), 2 (Albania), 3 (Algeria), 4 (American Samoa), 5 (Andorra), 6 (Angola), 7 (Anguilla), 8 (Antarctica), 9 (Antigua and Barbuda), 10 (Argentina), 11 (Armenia), 12 (Aruba), 13 (Australia), 14 (Austria), 15 (Azerbaijan), 16 (Baha
`estimate_only`	boolean		If true, do NOT create the audience — only estimate its size and return the expected contact count. Use this when the user wants to preview LinkedIn audience size before committing. Defaults to false (audience is created).

Full page, schema and recorded response · machine index

create_marketing_automation_contacts_audience Writes Create a Marketing-Automation-contacts custom audience (platform customAudienceType MA_CONTACTS_STATIC or MA_CONTACTS_DYNAMIC) from a provider library list.

Creates or changes something in the account.

Parameter	Type		Notes
`custom_audience_name`	string	required	Audience display name. Must not contain '/'.
`library_list_id`	integer	required	The library list id from list_marketing_automation_lists.
`provider`	string	required	The Marketing-Automation provider that owns the list. one of: `HUBSPOT`, `MARKETO`
`variant`	string	required	STATIC = snapshot (MA_CONTACTS_STATIC); DYNAMIC = refreshing (MA_CONTACTS_DYNAMIC). Ask the user if unspecified. one of: `STATIC`, `DYNAMIC`
`list_name`	string		Optional: the source list's own name, from list_marketing_automation_lists.

Full page, schema and recorded response · machine index

create_reddit_target_group Writes Create a new Reddit target group with name and targeting criteria (communities, interests, keywords, location).

Creates or changes something in the account.

Parameter	Type		Notes
`name`	string	required	The name for the new Reddit target group. Must be unique, descriptive, and UNDER 50 characters — the platform rejects 50+ character names with a 400.
`targeting`	object	required	Complete Reddit targeting configuration with include/exclude rules. LOCATION must be the first include block. All references must come from search_reddit_criteria.

Full page, schema and recorded response · machine index

create_retargeting_audience Writes Create a new retargeting audience on the Metadata platform from native channel audiences.

Creates or changes something in the account.

Parameter	Type		Notes
`audience_data`	object	required	Complete audience object from get_retargeting_audiences response. Should contain: id, name, type, subtype, channel, matchCount, externalAccountId, and other relevant fields.

Full page, schema and recorded response · machine index

create_salesforce_contacts_audience Writes Create a Salesforce-contacts custom audience (platform customAudienceType SFDC_CONTACTS_STATIC or SFDC_CONTACTS_DYNAMIC) from a Salesforce mixed-criteria.

Creates or changes something in the account.

Parameter	Type		Notes
`name`	string	required	Audience display name.
`criteria_id`	integer	required	The Salesforce mixed-criteria id from list_salesforce_mixed_criteria.
`variant`	string	required	STATIC = snapshot (SFDC_CONTACTS_STATIC); DYNAMIC = refreshing (SFDC_CONTACTS_DYNAMIC). Ask the user if unspecified. one of: `STATIC`, `DYNAMIC`

Full page, schema and recorded response · machine index

create_segment Writes Create a new segment with insights criteria.

Creates or changes something in the account.

Parameter	Type		Notes
`name`	string	required	Descriptive, unique segment name.
`criteria`	object	required	Criteria object with operator and match fields.

Full page, schema and recorded response · machine index

create_target_group Writes Create a new target group with name and targeting criteria.

Creates or changes something in the account.

Parameter	Type		Notes
`name`	string	required	The name for the new target group. Must be unique, descriptive, and UNDER 50 characters — the platform rejects 50+ character names with a 400.
`channel`	string	required	The advertising channel for the target group. Only LINKEDIN is currently supported. one of: `LINKEDIN`
`targeting`	object		Complete targeting configuration with include/exclude rules...

Full page, schema and recorded response · machine index

create_technographic_audience Writes Creates B2B audiences based on technology stack, software usage, and firmographic attributes.

Creates or changes something in the account.

Parameter	Type		Notes
`name`	string	required	Audience name (required). Maximum 50 characters.
`technologies`	array	required	Array of technology names to target (required). Example: ['Salesforce', 'HubSpot', 'Marketo']
`industries`	array		Industry IDs for firmographic/technographic filtering (optional). Valid IDs: 1=Defense & Space, 3=Computer Hardware, 4=Computer Software, 5=Computer Networking, 6=Internet, 7=Semiconductors, 8=Telecommunications, 9=Law Practice, 10=Legal Services, 11=Management Consulting, 12=Biotechnology, 13=Medic
`employees`	array		Employee count ranges for firmographic filtering (optional). Valid enum values: LESS_THAN_10 (0-9), FROM_10_TO_50 (10-49), FROM_50_TO_200 (50-199), FROM_200_TO_500 (200-499), FROM_500_TO_1000 (500-999), FROM_1000_TO_5000 (1,000-4,999), FROM_5000_TO_10000 (5,000-9,999), OVER_10000 (10,000+). Example:
`revenue`	array		Revenue ranges for firmographic filtering (optional). Valid enum values: LESS_THAN_1M ($0-$1M), FROM_1M_TO_10M ($1M-$10M), FROM_10M_TO_50M ($10M-$50M), FROM_50M_TO_100M ($50M-$100M), FROM_100M_TO_200M ($100M-$200M), FROM_200M_TO_500M ($200M-$500M), FROM_500M_TO_1000M ($500M-$1B), OVER_1000M ($1B+).
`location_country_ids`	array		Country IDs for geographic targeting (optional). Valid IDs: 1 (Afghanistan), 2 (Albania), 3 (Algeria), 4 (American Samoa), 5 (Andorra), 6 (Angola), 7 (Anguilla), 8 (Antarctica), 9 (Antigua and Barbuda), 10 (Argentina), 11 (Armenia), 12 (Aruba), 13 (Australia), 14 (Austria), 15 (Azerbaijan), 16 (Baha
`location_state_ids`	array		State IDs for geographic targeting (optional). Valid US State IDs: 1 (Alabama), 2 (Alaska), 3 (Arizona), 4 (Arkansas), 5 (California), 6 (Colorado), 7 (Connecticut), 8 (Delaware), 9 (Florida), 10 (Georgia), 11 (Hawaii), 12 (Idaho), 13 (Illinois), 14 (Indiana), 15 (Iowa), 16 (Kansas), 17 (Kentucky),
`job_title_includes`	array		Free text keywords for job titles to include (optional). Example: ['Software Engineer', 'Manager']
`job_title_excludes`	array		Free text keywords for job titles to exclude (optional). Example: ['Senior', 'Lead']
`job_function_include_ids`	array		Job function IDs to include (optional). Valid IDs: 1 (Marketing), 2 (Sales), 3 (Customer Service), 4 (Finance), 5 (HR), 6 (PR), 7 (Media), 8 (Design), 9 (Engineering), 10 (Operations), 11 (Legal), 12 (Education), 13 (Health), 14 (Real Estate), 15 (Accounting), 16 (Administrative), 17 (Business Devel
`job_function_exclude_ids`	array		Job function IDs to exclude (optional). Valid IDs: 1-28 as listed above
`seniority_include_ids`	array		Seniority level IDs to include (optional). Valid IDs: 2 (Manager), 3 (Director), 4 (VP), 5 (CXO), 6 (Unpaid), 7 (Training), 8 (Senior), 9 (Owner), 10 (Entry), 11 (Partner). Example: [10, 7] for Entry and Training
`seniority_exclude_ids`	array		Seniority level IDs to exclude (optional). Valid IDs: 2-11 as listed above
`contacts_per_company_limit`	integer		Maximum number of contacts to target per company (optional). Example: 20
`estimate_only`	boolean		If true, do NOT create the audience — only estimate its size and return the counts. Use this when the user wants to preview how many companies/contacts match the criteria before committing. Defaults to false (audience is created).

Full page, schema and recorded response · machine index

estimate_target_group Read Estimate audience size for target group criteria before creating.

Returns data. Safe to call freely.

Parameter	Type		Notes
`channel`	string	required	Channel to estimate for. LINKEDIN (default) takes criteria ids from search_target_group_criteria; REDDIT takes them from search_reddit_criteria and is estimated the same way (PRD-31759). Omit for LinkedIn. one of: `LINKEDIN`, `REDDIT`
`targeting`	object		Complete targeting configuration with include/exclude rules. Structure: { "include": [ { "criteria": [{"type": str, "data": object}], "audiences": [] } ], "exclude": {"criteria": [], "audiences": []} } All criteria IDs/references must come from search_target_group_criteria tool.

Full page, schema and recorded response · machine index

find_lookalike_companies Read Find companies similar to one or more seed domains (for audience expansion).

Returns data. Safe to call freely.

Parameter	Type		Notes
`domains`	array	required	Seed company domains, e.g. ['metadata.io'].
`limit`	integer		Max lookalikes to return (default 10).

Full page, schema and recorded response · machine index

get_abm_account_lists Read Get account lists from the Metadata ABM Lists library.

Returns data. Safe to call freely.

Parameter	Type	Notes
`filename`	string	Filter by account list filename (supports partial matching)
`page`	integer	Page number for pagination (0-based)
`size`	integer	Number of results per page
`sort`	string	Sort criteria (e.g., 'createdDate,desc', 'name,asc', 'accountsCount,desc')
`status`	string	Filter by status (e.g., READY, PROCESSING, ERROR)
`type`	string	Filter by list type (e.g., CSV, WEBSITE_RETARGETING)

Full page, schema and recorded response · machine index

get_audience_details Read Custom audience information from Metadata platform: configurations, targeting criteria, size metrics, performance.

Returns data. Safe to call freely.

Parameter	Type		Notes
`page`	integer		Page number for pagination (0-based)
`size`	integer		Number of results per page
`sort`	string		Sort criteria (e.g., 'creationDate,desc', 'name,asc', 'audienceSize,desc')
`name`	string	required	Filter by audience name (supports partial matching)

Full page, schema and recorded response · machine index

get_deep_audience_details Read Get detailed information about audiences including criteria and advanced configuration.

Returns data. Safe to call freely.

Parameter	Type		Notes
`audience_ids`	array	required	Array of audience IDs to retrieve detailed information for. Example: [42828, 42829, 42831]

Full page, schema and recorded response · machine index

get_intent_topics Read Get available intent topics for BOMBOORA audience creation.

Returns data. Safe to call freely.

Takes no arguments.

Full page, schema and recorded response · machine index

get_matched_audiences Read Get matched audiences (custom audiences) from the Campaign Wizard based on audience type and targeting options.

Returns data. Safe to call freely.

Parameter	Type		Notes
`type`	array	required	Array of audience targeting types to filter by. If omitted, searches all targeting types.
`channel`	string		The advertising channel to filter campaigns by one of: `LINKEDIN`, `FACEBOOK`, `GOOGLE_ADS`, `INSTAGRAM`, `TWITTER`, `YOUTUBE`, `REDDIT`, `MICROSOFT_ADS`
`page`	integer		Page number for pagination (0-indexed)
`sort`	string		Sort field and direction in format 'field,direction'. Common fields: createdDate. Direction: asc or desc.
`size`	integer		Number of results per page

Full page, schema and recorded response · machine index

get_retargeting_audiences Read Get all available retargeting audiences for a specific advertising channel.

Returns data. Safe to call freely.

Parameter	Type		Notes
`channel`	string	required	The advertising channel to fetch retargeting audiences for one of: `GOOGLE_ADS`, `LINKEDIN`, `FACEBOOK`

Full page, schema and recorded response · machine index

get_segment_criteria Read Get full criteria object for a segment by ID or name.

Returns data. Safe to call freely.

Parameter	Type		Notes
`segment_id`	integer		Segment ID (fastest lookup).
`segment_name`	string		Segment name (fuzzy matching, case-insensitive).

Full page, schema and recorded response · machine index

list_marketing_automation_lists Read List a Marketing-Automation provider's contact library lists — the selectable source for creating a Marketing-Automation-contacts audience.

Returns data. Safe to call freely.

Parameter	Type		Notes
`provider`	string	required	The Marketing-Automation provider to list lists for. one of: `HUBSPOT`, `MARKETO`
`list_type`	string		Optional list-type filter. If omitted, the provider's own list types are queried and merged (HubSpot: STATIC_LIST + ACTIVE_LIST; Marketo: all three). one of: `SMART_LIST`, `STATIC_LIST`, `ACTIVE_LIST`
`name`	string		Optional list-name filter (partial match).

Full page, schema and recorded response · machine index

list_salesforce_mixed_criteria Read List the account's saved Salesforce "mixed criteria" (contact-list criteria) — the selectable source for creating a Salesforce-contacts audience.

Returns data. Safe to call freely.

Parameter	Type		Notes
`qualified`	boolean		When true, keep only criteria whose generated filter matches qualified-signal fields.

Full page, schema and recorded response · machine index

list_segments Read List pre-saved segments (reusable filter configurations).

Returns data. Safe to call freely.

Parameter	Type		Notes
`name`	string		Optional fuzzy search by name.
`sort`	string		Sort format: field,direction. Default: name,asc

Full page, schema and recorded response · machine index

list_target_groups Read List and search target groups in Metadata platform.

Returns data. Safe to call freely.

Parameter	Type		Notes
`name`	string		Optional name filter for searching target groups (partial matching, case-insensitive). Example: 'engineer' matches 'Engineers EMEA', 'Senior Engineers', etc.
`sort`	string		Sort criteria. Options: 'createdDate,desc' (newest first), 'createdDate,asc' (oldest first), 'name,asc' (A-Z), 'name,desc' (Z-A). Default: 'createdDate,desc'

Full page, schema and recorded response · machine index

remove_audiences_from_campaign Destructive Remove specific audiences from one or more channels in an existing campaign.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`campaign_id`	integer	required	The unique identifier of the campaign
`channel_audiences`	object	required	Map of channel type to list of audience names to remove. Example: {"FACEBOOK": ["Audience A"], "LINKEDIN": ["Audience B"]}

Full page, schema and recorded response · machine index

remove_target_groups_from_campaign Destructive Remove targeting groups from one or more channels in an existing campaign.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`campaign_id`	integer	required	The unique identifier of the campaign
`channel_target_groups`	object	required	Map of channel type to list of targeting group names to remove. Example: {"LINKEDIN": ["Tech Decision Makers"]}

Full page, schema and recorded response · machine index

retrieve_target_group_by_id Read Get detailed target group information by ID.

Returns data. Safe to call freely.

Parameter	Type		Notes
`target_group_id`	integer	required	The unique ID of the target group to retrieve. Must be a positive integer. If you don't have the ID, use list_target_groups to search by name first.

Full page, schema and recorded response · machine index

search_reddit_criteria Read Get available Reddit targeting criteria and their reference values (externalId + name) for target group creation.

Returns data. Safe to call freely.

Parameter	Type		Notes
`criteria_type`	string	required	Type of Reddit criteria to search. communities/interests/keywords require search_terms; locations returns countries/states (search_terms optional). one of: `communities`, `interests`, `keywords`, `locations`
`search_terms`	array		Search keywords (min 2 characters). Required for communities/interests/keywords. Example: ['technology']

Full page, schema and recorded response · machine index

search_target_group_criteria Read Get available criteria and their supported values (names and IDs) for target group creation/updates.

Returns data. Safe to call freely.

Parameter	Type		Notes
`channel`	string	required	The advertising channel for which to retrieve criteria. Only LINKEDIN is supported. one of: `LINKEDIN`
`criteria_type`	string		Type of targeting criteria to retrieve or search. List-based (returns predefined options): - age-ranges: Age range segments - company-categories: Company type classifications - company-growth-rates: Company growth classifications - revenues: Company revenue ranges - employees: Employee count ranges one of: `age-ranges`, `company-categories`, `company-growth-rates`, `revenues`, `employees`, `industry-taxonomy`, `jobFunctions`, `seniority`, `followed-companies`, `locations`, `job-title`, `member-groups`
`search_terms`	array		Array of search keywords for search-based criteria (job-title, member-groups, interests, traits). Example: ['engineer', 'developer']
`reference_type`	string		Specific reference type for search operations (LinkedIn only). Auto-determined from criteria_type and channel if not provided. one of: `LINKEDIN_JOB_TITLES`, `LINKEDIN_MEMBER_GROUPS`, `BUYER_INTENT_INDUSTRIES`
`exact_match`	boolean		Whether to require exact matches in search (default: true). Set to false for broader/fuzzy matching.

Full page, schema and recorded response · machine index

unarchive_audience Writes Unarchive (restore) a previously archived custom audience on the Metadata platform.

Creates or changes something in the account.

Parameter	Type		Notes
`custom_audience_id`	integer		The numeric ID of the custom audience to unarchive.
`audience_name`	string		The name of the custom audience to unarchive. If provided without custom_audience_id, the tool will search for the audience by name to resolve the ID.

Full page, schema and recorded response · machine index

update_target_group Destructive Update existing target group with new name and/or targeting criteria.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`target_group_id`	integer	required	The unique ID of the target group to update. Must be a positive integer. Use list_target_groups to find the ID by name.
`name`	string	required	REQUIRED: Target group name. CRITICAL - This is a PUT request so you MUST provide the name even if not changing it. Use retrieve_target_group_by_id to get the current name if you don't have it.
`targeting`	object		Updated targeting configuration (optional). Structure must match create_target_group format. Only update criteria that have isModifiable=true in the current configuration. All criteria IDs/references must come from search_target_group_criteria tool. REQUIRED: Must estimate this targeting first and v

Full page, schema and recorded response · machine index

upload_account_list_csv_audience Writes Upload a list of company ACCOUNTS (name + website) as an Account List CSV and create a FIRMOGRAPHIC_INCLUDE audience on the Metadata platform, optionally overlaid with platform-…

Creates or changes something in the account.

Parameter	Type		Notes
`audience_name`	string	required	Name for the new audience (required).
`companies`	object		Inline map of company names to website URLs (optional). Example: {"Acme Corp": "https://acme.com", "Globex": "https://globex.com"}. MUTUALLY EXCLUSIVE with `companies_source_csv_url` — provide exactly one.
`companies_source_csv_url`	string		Public URL of a CSV file with header EXACTLY `companyname,companywebsite` (case-insensitive). The MCP server downloads, validates, and parses the file. Use this when the user attached a CSV to the chat — the URL is surfaced via `AudienceBrief.attached_file_urls`. MUTUALLY EXCLUSIVE with `companies`.
`location_country_ids`	array		Country IDs for geographic targeting (optional). Valid IDs: 1 (Afghanistan), 2 (Albania), 3 (Algeria), 4 (American Samoa), 5 (Andorra), 6 (Angola), 7 (Anguilla), 8 (Antarctica), 9 (Antigua and Barbuda), 10 (Argentina), 11 (Armenia), 12 (Aruba), 13 (Australia), 14 (Austria), 15 (Azerbaijan), 16 (Baha
`location_state_ids`	array		State IDs for geographic targeting (optional). Valid US State IDs: 1 (Alabama), 2 (Alaska), 3 (Arizona), 4 (Arkansas), 5 (California), 6 (Colorado), 7 (Connecticut), 8 (Delaware), 9 (Florida), 10 (Georgia), 11 (Hawaii), 12 (Idaho), 13 (Illinois), 14 (Indiana), 15 (Iowa), 16 (Kansas), 17 (Kentucky),
`job_title_includes`	array		Free text keywords for job titles to include (optional). Example: ['Software Engineer', 'Manager']
`job_title_excludes`	array		Free text keywords for job titles to exclude (optional). Example: ['Senior', 'Lead']
`job_function_include_ids`	array		Job function IDs to include (optional). Valid IDs: 1 (Marketing), 2 (Sales), 3 (Customer Service), 4 (Finance), 5 (HR), 6 (PR), 7 (Media), 8 (Design), 9 (Engineering), 10 (Operations), 11 (Legal), 12 (Education), 13 (Health), 14 (Real Estate), 15 (Accounting), 16 (Administrative), 17 (Business Devel
`job_function_exclude_ids`	array		Job function IDs to exclude (optional). Valid IDs: 1-28 as listed above
`seniority_include_ids`	array		Seniority level IDs to include (optional). Valid IDs: 2 (Manager), 3 (Director), 4 (VP), 5 (CXO), 6 (Unpaid), 7 (Training), 8 (Senior), 9 (Owner), 10 (Entry), 11 (Partner). Example: [10, 7] for Entry and Training
`seniority_exclude_ids`	array		Seniority level IDs to exclude (optional). Valid IDs: 2-11 as listed above
`contacts_per_company_limit`	integer		Maximum number of contacts to target per company (optional). Example: 20

Full page, schema and recorded response · machine index

upload_contact_list_csv_audience Writes Upload a list of individual CONTACTS (people) as a Contact List CSV and create a CONTACT_LIST audience on the Metadata platform.

Creates or changes something in the account.

Parameter	Type		Notes
`audience_name`	string	required	Name for the new contact list audience (required).
`contacts`	array		Inline array of contact objects (optional). Each object represents one person/contact with fields: email (required), firstname, lastname, jobtitle, employeecompany, country, appleidfa, googleaid. Minimum 300 contacts, maximum 300,000 contacts. MUTUALLY EXCLUSIVE with `contacts_source_csv_url` — prov
`contacts_source_csv_url`	string		Public URL of a contacts CSV file with header EXACTLY `email,firstname,lastname,jobtitle,employeecompany,country,appleidfa,googleaid` (case-insensitive). The MCP server downloads, validates, and uploads the file. Use this when the user attached a CSV to the chat — the URL is surfaced via `AudienceBr

Full page, schema and recorded response · machine index

add_and_edit_campaign_elements Destructive Add new elements and edit existing properties of a **Precision Optimization (1×1×1 / METADATA)** marketing campaign.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`campaign_data`	object	required	Campaign data with fields to edit and/or elements to add

Full page, schema and recorded response · machine index

add_and_edit_native_campaign_elements Destructive Add elements to, and edit campaign-level fields of, an existing **Channel-First / Native (N×N×N)** campaign — the one built with `create_native_structure_campaign` (channel.stru…

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`campaign_data`	object	required	Native (N×N×N) campaign edit payload: campaignId + optional campaign-level fields + per-channel `containers[]` (append) and optional `target_groups[]`.

Full page, schema and recorded response · machine index

check_campaign_launch_readiness Read Check whether a draft campaign is ready to launch, per channel.

Returns data. Safe to call freely.

Parameter	Type		Notes
`campaign_id`	integer	required	The wizard campaign ID to check launch readiness for

Full page, schema and recorded response · machine index

connect_channel Writes Get the OAuth redirect URL to connect a specific advertising channel to the Metadata platform.

Creates or changes something in the account.

Parameter	Type		Notes
`channel`	string	required	The channel to connect. Must be one of: linkedin, meta (for Facebook/Instagram), google, reddit, x (for Twitter) one of: `linkedin`, `meta`, `google`, `reddit`, `x`, `facebook`, `instagram`, `twitter`
`acknowledge_data_access`	boolean		Set true to confirm the person you are acting for asked to connect this channel. The response is an OAuth authorization URL that starts a credential grant into their Metadata account. Without this the call is refused.

Full page, schema and recorded response · machine index

create_budget_group Writes Create or update advertising budget and spending allocation.

Creates or changes something in the account.

Parameter	Type		Notes
`data`	object	required	Budget group configuration

Full page, schema and recorded response · machine index

create_campaign Writes Create multi-channel advertising campaigns for LinkedIn, Facebook, Instagram, Google Ads, Reddit, and Microsoft Ads using the platform's "Precision Optimization" (1×1×1) model —…

Creates or changes something in the account.

Parameter	Type		Notes
`campaign_data`	object	required	Complete campaign configuration data

Full page, schema and recorded response · machine index

create_native_structure_campaign Writes Create a multi-channel campaign using the platform's "Channel First Structure" (N×N×N) — the UI also labels it "Native Structure" / "Platform-First Structure".

Creates or changes something in the account.

Parameter	Type		Notes
`campaign_data`	object	required	Complete Channel First / Native Structure (N×N×N) campaign configuration. Each channel takes a `containers` list; each container groups one audience with an `ad_offers` list of {ad, offer} pairs.

Full page, schema and recorded response · machine index

create_update_offer Destructive Create or update Lead Gen or Landing Page offers in Metadata platform.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`offer_type`	string	required	Type of offer to create one of: `Lead Gen`, `Landing Page`
`data`	object	required	Offer configuration data. Required fields depend on offer_type and channel - see description for details.

Full page, schema and recorded response · machine index

disconnect_channel Destructive Disconnect a specific advertising channel from the Metadata platform.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`channel`	string	required	The channel to disconnect. Must be one of: reddit, facebook, linkedin, twitter, google, instagram one of: `reddit`, `facebook`, `linkedin`, `twitter`, `google`, `instagram`

Full page, schema and recorded response · machine index

duplicate_campaign Writes Duplicate (copy) an existing campaign into a NEW DRAFT campaign.

Creates or changes something in the account.

Parameter	Type		Notes
`campaign_id`	integer	required	The wizard/campaign ID of the campaign to duplicate. Must be in status Launched, Launching, Editing, or Draft.
`name`	string		Name for the new (copied) campaign. Must be unique on the account and <= 255 chars. Omit to auto-generate '<original name>_copy_<timestamp>'.

Full page, schema and recorded response · machine index

find_offer_url Read Locate the BEST specific landing-page URL for a marketing offer on a given domain.

Returns data. Safe to call freely.

Parameter	Type		Notes
`domain`	string	required	Company website URL or domain. Examples: 'snowflake.com', 'www.servicenow.com'
`offer_type`	string	required	Type of offer content to locate. Drives which paths are probed and which keywords score higher. one of: `case_study`, `demo`, `whitepaper`, `roi_calculator`, `guide`, `report`, `ebook`, `webinar`
`topic`	string		Optional topic keywords to bias scoring (e.g., 'fintech', 'servicenow integration', 'data warehousing'). Improves hit rate when the domain has many offers of the same type.

Full page, schema and recorded response · machine index

get_budget_group Read Get configuration and settings for a budget group (NOT performance data).

Returns data. Safe to call freely.

Parameter	Type		Notes
`name`	string	required	The name of the budget group to search for. Example: 'Brand'

Full page, schema and recorded response · machine index

get_campaign_by_wizard_id Read Get full campaign details by its wizard campaign ID, including channels, audiences, ads, and budget configuration.

Returns data. Safe to call freely.

Parameter	Type		Notes
`campaign_id`	integer	required	The wizard campaign ID to retrieve details for

Full page, schema and recorded response · machine index

get_campaign_targeting_locations Read Where a campaign actually TARGETS, by channel, with the platform's location ids resolved to country and state names.

Returns data. Safe to call freely.

Parameter	Type	Notes
`campaign_ids`	array	Wizard campaign ids to read. Omit to scan the account.
`country_names`	array	Filter to campaigns targeting these countries, e.g. ["Germany", "Austria"].
`max_campaigns`	integer	Cap on campaigns read when scanning (default 25, max 100).

Full page, schema and recorded response · machine index

get_offer Read Get offer information by name or list all offers.

Returns data. Safe to call freely.

Parameter	Type		Notes
`name`	string		The name of the offer to search for. Example: 'AI Marketing Audit Demo'. If omitted, returns a list of all available offers.
`offer_id`	integer		Exact offer id. PREFER THIS over `name` whenever you have an id: a name lookup returns the first search match, so offers sharing a prefix (a template name plus a random suffix, which is how offers get named here) can resolve to the wrong one silently. Takes precedence when both are given.

Full page, schema and recorded response · machine index

launch_campaign Destructive LAUNCHES REAL MARKETING CAMPAIGNS WITH REAL BUDGET SPEND 🚨 KEYWORDS: launch, start, activate, go live, campaign, begin, advertising, spend, marketing PREREQUISITE: Before runni…

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`campaign_id`	integer	required	The unique identifier of the draft campaign to launch

Full page, schema and recorded response · machine index

list_budget_groups Read Browse and list all available budget groups in the Metadata platform with customizable sorting.

Returns data. Safe to call freely.

Parameter	Type	Notes
`sort_by`	string	Field to sort budget groups by. Default: totalActiveExperiments one of: `totalActiveExperiments`, `monthlyCap`, `averageDailyBudgetPerExperiment`, `spent`, `totalBudget`, `endDate`, `startDate`, `leads`, `cpl`
`sort_order`	string	Sort order - 'asc' (ascending) or 'desc' (descending). Default: desc one of: `asc`, `desc`
`page`	integer	Page number for pagination (0-based). Default: 0
`size`	integer	Number of results per page. Default: 10

Full page, schema and recorded response · machine index

list_wizard_campaigns Read List wizard (Metadata-native) campaigns for the current account with pagination, sorting, and experiment-status filtering.

Returns data. Safe to call freely.

Parameter	Type	Notes
`page`	integer	0-based page index. Default: 0
`size`	integer	Results per page. Default: 25
`sort_by`	string	Field to sort by. Default: createdDate one of: `createdDate`, `startDate`, `endDate`, `name`, `budget`, `spent`, `leads`, `cpl`, `cpc`, `ctr`, `impressions`, `clicks`
`sort_order`	string	Sort order. Default: desc one of: `asc`, `desc`
`status`	string	Campaign-level status filter. Comma-separated subset of Draft, Launched, Finished. Omit to include all three. Combine with experiment_statuses to mirror the UI tabs (e.g. Active tab = status=Launched + experiment_statuses=Active,WithoutSpend).
`experiment_statuses`	string	Experiment-level status filter. Comma-separated subset of Active, WithoutSpend, Paused, Completed, Failed. Omit to include all five.
`visibility`	string	Optional: filter by campaign visibility. Omit to include all. one of: `VISIBLE`, `ARCHIVED`
`name`	string	Optional: server-side name substring filter.

Full page, schema and recorded response · machine index

manage_campaign Destructive Manage a campaign by pausing or restarting it.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`campaign_id`	integer	required	The unique identifier of the campaign to manage
`action`	string	required	The action to perform: 'pause' to stop the campaign, 'restart' to resume it one of: `pause`, `restart`

Full page, schema and recorded response · machine index

manage_experiment Destructive Pause or restart a whole experiment (a launched channel campaign inside a wizard campaign).

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`experiment_id`	integer	required	The experiment ID to pause or restart (experimentId from search_experiments).
`action`	string	required	The action to perform: 'pause' to stop the experiment, 'restart' to resume it one of: `pause`, `restart`
`confirm_budget_impact`	boolean		Restart only. Set true ONLY after the user explicitly confirmed the budget-group impact returned by a previous manage_experiment call with status='requires_confirmation'.
`auto_pause_ignore_historical_data`	boolean		Restart only. When the experiment was auto-paused for low performance, true tells the optimizer to ignore the historical performance data that caused the pause.

Full page, schema and recorded response · machine index

remove_channels_from_campaign Destructive Disable one or more entire channels from an existing campaign.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`campaign_id`	integer	required	The unique identifier of the campaign
`channels`	array	required	List of channels to disable from the campaign. Example: ["FACEBOOK", "GOOGLE_ADS"]

Full page, schema and recorded response · machine index

remove_offers_from_campaign Destructive Remove offers — Lead Gen forms OR Landing Pages — from one or more channels in an existing campaign.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`campaign_id`	integer	required	The unique identifier of the campaign
`channel_offers`	object	required	Map of channel type to list of offer names (Lead Gen form or Landing Page) to remove. Example: {"FACEBOOK": ["Demo Form"], "GOOGLE_ADS": ["Pricing LP"]}

Full page, schema and recorded response · machine index

search_campaigns_by_names Read Search for campaigns by name to get campaign details, IDs, and status.

Returns data. Safe to call freely.

Parameter	Type		Notes
`campaign_names`	array	required	Array of campaign names or partial names to search for (case-insensitive, partial matching)

Full page, schema and recorded response · machine index

search_experiment_recipes Read Find proven CAMPAIGN CONFIGURATIONS to build from — combinations of channel, ad format, audience type, bid strategy and budget band that have run together successfully across mu…

Returns data. Safe to call freely.

Parameter	Type		Notes
`query`	string	required	Natural-language description of the campaign you are building, e.g. "lead gen campaign targeting IT decision makers", "top of funnel awareness push for a new product launch".
`channel`	any		Optional exact filter — "linkedin", "facebook", "google_ads", "instagram". Omit to search every channel.
`ad_format`	any		Optional exact filter — "image", "video", "document", "carousel", "text".
`audience_type`	any		Optional exact filter, e.g. "firmographic_include", "linkedin_retargeting", "native_targeting_dynamic".
`bid_strategy`	any		Optional exact filter, e.g. "auto_bid", "manual_bid", "metadata_bid_agent".
`num_results`	integer		How many recipes to return (default 5).

Full page, schema and recorded response · machine index

search_experiments Read Search for experiments by name, status, and budget group with flexible filtering.

Returns data. Safe to call freely.

Parameter	Type	Notes
`search_name`	string	The experiment or campaign name to search for (partial, case-insensitive matching). Optional - omit to skip name filtering.
`budget_group_ids`	array	Optional array of budget group IDs to filter experiments. Only returns experiments that belong to these budget groups.
`launched_experiment_statuses`	array	Optional array of experiment statuses to filter by. Valid values: Active, WithoutSpend, Paused, Completed, Failed, Disconnected
`page`	integer	Page number for pagination (default: 0)
`size`	integer	Number of results per page (default: 5)
`sort`	string	Sort criteria in format 'field,direction' (e.g., 'name,asc')

Full page, schema and recorded response · machine index

update_experiments_daily_budgets Destructive Update daily budget allocations for one or more experiments in the Metadata platform.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`budgets`	object	required	Dictionary mapping experiment IDs to new daily budget values. Keys are experiment IDs (integers or strings), values are budget amounts (numbers or strings with comma/dot separators). Example: {'12345': 100.50, '12346': '150,75'}

Full page, schema and recorded response · machine index

update_reddit_channel_settings Destructive Update the Reddit channel settings of an existing DRAFT wizard campaign: native objective, allow comments, placements (feed/conversations), conversion pixel, and (for manual-bid…

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`campaign_id`	integer	required	The unique identifier of the wizard campaign
`objective`	string		Reddit native objective. IMPRESSIONS = Brand awareness and Reach; CLICKS = Traffic (the only objective allowed on Lead Gen campaigns); VIDEO_VIEWABLE_IMPRESSIONS_6S = Video Views 6s (video-only channels). one of: `IMPRESSIONS`, `CLICKS`, `VIDEO_VIEWABLE_IMPRESSIONS_6S`
`allow_comments`	boolean		Allow Redditors to comment on the promoted post. Platform default is true.
`placements`	object		Where the ads serve. At least one placement must stay selected; conversations is forced off when the channel has video ads.
`conversion_pixel_id`	string		Reddit conversion pixel id (created in Reddit Ads Manager). Empty string clears the pixel.
`bid_amount`	number		Manual bid amount, applied only when the effective objective uses MANUAL_BID (IMPRESSIONS / Video Views).

Full page, schema and recorded response · machine index

archive_ad Destructive Archive (soft-delete) one or more ads on the Metadata platform.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`ids`	array	required	Ad id(s) to archive. At least one integer required; pass multiple to archive in bulk.

Full page, schema and recorded response · machine index

create_update_carousel_ad Destructive Create OR update a CAROUSEL ad on Reddit (Reddit only).

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`adName`	string	required	Ad name. Required, max 50 chars.
`headline`	string	required	Headline shared across all cards. Required, max 300.
`cards`	array	required	2 to 6 carousel cards. Each card needs a distinct image (no duplicates).
`text`	string		Optional body text. Max 2000.
`thumbnailLibraryId`	integer		Optional thumbnail image library id.
`completionStatus`	string		DRAFT (default) persists locally without publishing; COMPLETED finalizes the ad. one of: `DRAFT`, `COMPLETED`
`id`	integer		Include to UPDATE an existing carousel ad; omit to CREATE.

Full page, schema and recorded response · machine index

create_update_convo_ad Destructive Create or update a LinkedIn Conversation Ad ("convo ad") — a branching in-message chat where each step contains a message bubble and a list of button actions that drive the conv…

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`id`	integer		Existing convo ad ID — include to UPDATE; omit to CREATE.
`name`	string	required	Ad name in the platform library (max 50 chars).
`senderId`	string	required	LinkedIn sender ID. MUST come from get_linkedin_senders.
`senderName`	string	required	Display name matching senderId from get_linkedin_senders.
`headlineText`	string	required	Subject line shown above the message thread.
`bannerCreativeLibraryId`	integer		Optional: imageLibraryId of the banner image. Upload first via upload_image_creative.
`completionStatus`	string		Defaults to DRAFT. one of: `DRAFT`, `COMPLETED`
`flowId`	integer		LinkedIn convo flow version. Defaults to 7 — keep the default unless told otherwise.
`templateId`	integer		LinkedIn convo template ID. Defaults to 1 — keep the default unless told otherwise.
`steps`	array	required	Ordered list of conversation steps. MUST include stepId=1 as the entry.

Full page, schema and recorded response · machine index

create_update_document_ad Destructive Create a new or update an existing LinkedIn Document Ad on the Metadata platform.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`id`	integer		Existing Document Ad ID — include to UPDATE; omit to CREATE.
`name`	string	required	Ad name in the platform library (max 50 characters).
`libraryId`	integer	required	Integer id of the DOCUMENT creative in the library (contentType=DOCUMENT). Get it from search_library_creatives_by_name(contentTypes="DOCUMENT"). NOT an image/video id.
`headline`	string	required	Headline shown with the document (max 200 characters).
`text`	string	required	Introductory text shown with the document (max 3000 characters).
`ctaType`	string		Call-to-action. Defaults to UNLOCK_FULL_DOCUMENT (the document-gate CTA). one of: `UNLOCK_FULL_DOCUMENT`
`maxPreviewPages`	integer		How many document pages are previewable before the unlock gate. Defaults to 1.
`completionStatus`	string		Defaults to DRAFT. Use COMPLETED only when every required field is final. one of: `DRAFT`, `COMPLETED`

Full page, schema and recorded response · machine index

create_update_google_ads_ad Destructive Create a new or update an existing Google Ads text ad on the Metadata platform.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`id`	integer		The existing ad ID to update. If omitted, a new ad will be created.
`name`	string	required	The name/label for the ad in Metadata platform (max 50 characters)
`headlines`	array	required	List of headline texts for the Google Ads ad (minimum 3, maximum 15 headlines, each max 30 characters)
`descriptions`	array	required	List of description texts for the Google Ads ad (2 to 4 descriptions). CRITICAL #1 FAILURE MODE: each description MUST be ≤ 90 characters — this is enforced by maxLength schema validation AND by server-side truncation. Going over 90 means either (a) the call is rejected before it runs, or (b) the de

Full page, schema and recorded response · machine index

create_update_image_ad Destructive Create OR update an IMAGE (or GIF) ad on Facebook, Instagram, LinkedIn, and/or Reddit.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`ads_data`	array	required	Array of IMAGE ads to create or update. Each item targets one or more of FB/IG/LinkedIn/Reddit; the platform creates one ad row per (ad, channel).

Full page, schema and recorded response · machine index

create_update_linkedin_message_ad Destructive Create a new or update an existing LinkedIn Message Ad on the Metadata platform.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`id`	integer		Existing Message Ad ID — include to UPDATE; omit to CREATE.
`name`	string	required	Ad name in the platform library (max 50 characters).
`senderId`	string	required	LinkedIn sender ID. MUST come from get_linkedin_senders. Do not invent.
`senderName`	string	required	Display name matching senderId, also from get_linkedin_senders.
`headlineText`	string	required	Message subject line shown in the recipient's inbox (max 60 characters).
`messageText`	string	required	Message body (max 2000 characters). Supports LinkedIn personalization macros — ONLY %FIRSTNAME%, %LASTNAME%, %COMPANYNAME%, %JOBTITLE%, %INDUSTRY%. Any other %...% token is rejected by the platform validator.
`actionText`	string	required	Call-to-action button label (max 20 characters).
`bannerCreativeLibraryId`	integer		Optional: imageLibraryId of the banner image shown above the message. Upload the image first via upload_image_creative and pass the returned ID. Omit to skip the banner.
`completionStatus`	string		Defaults to DRAFT. Use COMPLETED only when every required field is final. one of: `DRAFT`, `COMPLETED`

Full page, schema and recorded response · machine index

create_update_microsoft_ads_ad Destructive Create a new or update an existing Microsoft Ads (Bing) text ad on the Metadata platform.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`id`	integer		The existing ad ID to update. If omitted, a new ad will be created.
`name`	string	required	The name/label for the ad in Metadata platform (max 50 characters)
`headlines`	array	required	List of headline texts for the Microsoft Ads ad (minimum 3, maximum 15 headlines, each max 30 characters). Every headline must be DISTINCT — the platform rejects the ad if two share the same text — and must not contain '!'.
`descriptions`	array	required	List of description texts for the Microsoft Ads ad (2 to 4 descriptions — 2 is the platform minimum, a single-description ad is rejected). Each description MUST be 90 characters or fewer and MUST be distinct from the others. Count every character (incl. spaces/punctuation/emoji); target 70-85 chars

Full page, schema and recorded response · machine index

create_update_video_ad Destructive Create OR update a VIDEO ad on Facebook, Instagram, LinkedIn, and/or Reddit.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`ads_data`	array	required	Array of VIDEO ads to create or update. Each item targets one or more of FB/IG/LinkedIn/Reddit; the platform creates one ad row per (ad, channel).

Full page, schema and recorded response · machine index

edit_brand_creative Destructive Edit existing creative image using AI-powered editing while preserving brand consistency.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`image_source`	string	required	URL of the creative image to edit. Must be a valid accessible image URL (PNG, JPG, WebP). Can be a URL returned by a previous generation or edit.
`edit_instruction`	string	required	Natural language description of the desired edit(s). Can include multiple changes in one request. Examples: 'Replace the robot with a dashboard', 'Change headline to BOOST ROI and make CTA say GET STARTED', 'Add a +50% ROI badge in the corner'
`domain`	string	required	Company domain for brand context. Used to fetch brand DNA for consistency. Examples: 'metadata.io', 'bloomreach.com'

Full page, schema and recorded response · machine index

fetch_creative_details Read Get detailed information for specific creative assets by their IDs.

Returns data. Safe to call freely.

Parameter	Type		Notes
`creative_id`	integer		creative id to be query

Full page, schema and recorded response · machine index

fix_flexible_creative Destructive Apply a free-text AI edit to a flexible creative's BACKGROUND image and return the edited image URL.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`image_source`	string	required	URL of the background image to edit (PNG/JPG/WebP). Typically the flexible creative's current background.
`edit_instruction`	string	required	Natural-language description of the fix(es). Can combine several changes in one request.
`domain`	string	required	Company domain for brand context (e.g. 'metadata.io').

Full page, schema and recorded response · machine index

generate_brand_creative Writes Generate branded advertising images and designs using AI and company Brand DNA.

Creates or changes something in the account.

Parameter	Type		Notes
`domain`	string	required	Company domain to extract brand DNA from. Examples: 'docebo.com', 'metadata.io', 'bloomreach.com'. Do not include protocol (http/https).
`headline`	string	required	Primary headline text to display in the creative. Keep to 25 characters or fewer for optimal rendering (hard cap 40 — the image model truncates or garbles longer text). Used to select aligned reference ads and guide visual hierarchy. Examples: 'Boost ROI by 30%', 'AI-Powered Marketing', 'Drive More
`instructions`	string	required	Natural language instructions describing the desired creative. Will be enhanced with brand DNA patterns including proven layouts, CTAs, colors, and tones. Examples: 'Create a professional B2B ad with clean design', 'Generate announcement for new AI feature launch', 'Design case study visual with mod
`platform`	string		Target platform for the creative. Each platform has specific dimensions, text limits, and best practices that will be automatically applied. Default: 'linkedin' one of: `linkedin`, `facebook`, `instagram`, `display`, `general`
`include_logo`	boolean		Whether to include a logo overlay on the generated creative. Set to false to get the raw creative without any logo. Default: true
`include_cta`	boolean		Whether to render a call-to-action (button or styled text link) in the creative. Set to false for surfaces that must not contain a CTA, e.g. stills that seed video ads, where the channel's ad unit provides the real CTA. Default: true
`logo_position`	string		Force a specific logo position instead of using AI-recommended placement. If not provided, an AI model analyzes the creative to find the optimal position. one of: `top-left`, `top-right`, `bottom-left`, `bottom-right`
`logo_size_percent`	number		Logo width as a percentage of the image width (e.g., 15 means the logo occupies 15% of the image width). If not provided, the AI recommends an optimal size. Typical range: 10-30.
`logo_url`	string		Optional direct URL to a logo image (e.g. the URL of a logo the user just uploaded in chat). When set, THIS logo is overlaid for this creative instead of the brand kit's logo or the one from Brandfetch. Use it when the user says things like 'use this logo' or 'use the logo I uploaded'. The uploaded
`use_website_background`	boolean		Use the brand website's background color as the creative's dominant background; brand primary colors then appear as accents (CTA button, shapes, highlights). Default: true. Set false when the user wants a brand-colored or model-chosen background instead. Silently ignored when the brand kit has no re

Full page, schema and recorded response · machine index

generate_brand_kit Writes Analyze a company's brand from their website and extract brand identity.

Creates or changes something in the account.

Parameter	Type		Notes
`domain`	string	required	Company website URL or domain
`brand_name`	string		Optional brand name override
`sample_ad_urls`	array		Optional ad image URLs to analyze (max 5)
`force_regenerate`	boolean		If True, regenerate brand kit even if one already exists. Default: False

Full page, schema and recorded response · machine index

generate_flexible_brand_creative Writes Generate an EDITABLE, layered brand creative: a clean AI background scene (no baked text) plus AI-placed, editable text/CTA/logo layers.

Creates or changes something in the account.

Parameter	Type		Notes
`domain`	string	required	Company domain for brand DNA (e.g. 'metadata.io'). No protocol.
`instructions`	string	required	Natural-language direction for the background scene and overall creative (e.g. 'clean B2B scene for an AI analytics launch').
`platform`	string		Target platform (sets the default aspect ratio). Default: 'linkedin'. one of: `linkedin`, `facebook`, `instagram`, `reddit`, `display`, `general`
`aspect_ratio`	string		Canvas shape, overriding the one 'platform' implies. Pass this whenever the user names a size or shape: '1:1' for square (e.g. 1080x1080), '16:9' for wide/landscape, '9:16' for a full-height story, '4:5' for portrait. Omit to keep the platform's default shape. one of: `1:1`, `16:9`, `9:16`, `4:5`
`headline`	string		Headline copy to place as an editable layer.
`subheadline`	string		Optional subheadline copy.
`body`	string		Optional supporting body copy. Pass ONLY when the user explicitly asked for body copy — extra layers crowd the layout.
`cta`	string		Optional call-to-action button text (e.g. 'Learn more').
`disclaimer`	string		Optional fine-print/disclaimer copy. Pass ONLY when the user explicitly asked for a disclaimer.
`include_logo`	boolean		Add a brand logo layer (resolved from the brand kit). Default: true.

Full page, schema and recorded response · machine index

get_ad_details Read Get complete details and configuration for one or more ads from the Metadata platform.

Returns data. Safe to call freely.

Parameter	Type		Notes
`ids`	object	required	Ad ID(s) to get details for. Can be: single integer (182124) or comma-separated string ('182124,182123').

Full page, schema and recorded response · machine index

get_brand_kit Read Retrieve current brand kit configuration for a company.

Returns data. Safe to call freely.

Parameter	Type		Notes
`domain`	string	required	Company website URL or domain. Examples: 'metadata.io', 'www.bloomreach.com', 'https://example.com'

Full page, schema and recorded response · machine index

regenerate_flexible_background Destructive Regenerate ONLY the background scene of a flexible creative, keeping the editable text/logo layers untouched.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`domain`	string	required	Company domain for brand DNA (e.g. 'metadata.io'). No protocol.
`instructions`	string	required	Direction for the new background scene.
`platform`	string		Target platform (sets the default aspect ratio). Default: 'linkedin'. one of: `linkedin`, `facebook`, `instagram`, `reddit`, `display`, `general`
`aspect_ratio`	string		Pass the creative's own canvas.aspect_ratio so the new background keeps its shape. Omitting it on a non-default canvas reshapes the creative and the existing layers no longer fit. one of: `1:1`, `16:9`, `9:16`, `4:5`
`image_source`	string		Optional URL of the current background to guide a same-style regeneration (image-to-image). Omit for a fresh scene.
`background_hex`	string		Optional 6-digit hex to pin the dominant background color (e.g. '#0A2540').

Full page, schema and recorded response · machine index

remove_ad_from_campaign Destructive Remove specific ads (ad groups) from one or more channels in an existing campaign.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`campaign_id`	integer	required	The unique identifier of the campaign
`channel_ads`	object	required	Map of channel type to list of ad names to remove. Example: {"LINKEDIN": ["Ad A"], "FACEBOOK": ["Ad B"]}

Full page, schema and recorded response · machine index

render_flexible_creative Writes Flatten a layered flexible-creative document into a final PNG (rendered from the SAME layout the editor shows) and upload it to the creatives library.

Creates or changes something in the account.

Parameter	Type		Notes
`design`	object	required	The layered creative document to flatten (canvas, background, layers, brand) — the same JSON the editor edits.
`name`	string		Optional library name for the rendered image. Omit it and one is derived from the doc as `<brand-domain>-<channel>-<width>x<height>.png`, which is normally what you want. Must end in a file extension.
`upload_to_library`	boolean		Push the rendered PNG to the creatives library. Default: true.
`archive_library_image_id`	string		Optional. The library image id this save supersedes (the creative's previous render). Archived on the platform ONLY after the new upload succeeds, so re-saving an editable creative retires the old PNG instead of piling up dead library images. Never affects an ad already built from it.

Full page, schema and recorded response · machine index

search_ads_by_names Read Search for ads by name in the Metadata platform library, with pagination.

Returns data. Safe to call freely.

Parameter	Type		Notes
`ad_names`	array	required	Array of ad names or partial names to search for (case-insensitive, partial matching)
`page`	integer		Zero-based page index (default: 0). Page numbering starts at 0.
`size`	integer		Results per page (default: 25, recommended max: 25). Larger windows blow LLM token budgets — only request more when the caller actually needs it.
`sort`	string		Spring sort clause, e.g. 'createdDate,desc' (default), 'id,desc', or 'name,asc'.
`status`	string		Which status bucket to return. 'active' (default) is the prior active-only behaviour. 'archived' returns ONLY archived (soft-deleted) ads, which is what you need to find an ad's id for unarchive_ad. 'all' returns both active and archived in one response. Maps to the platform's status query param sem one of: `active`, `archived`, `all`

Full page, schema and recorded response · machine index

search_library_creatives_by_name Read Search or list creative assets in the library by name, across ALL content types — images, videos, GIFs and documents.

Returns data. Safe to call freely.

Parameter	Type	Notes
`creativeName`	string	Name or partial name of the creative to search for. Supports partial matching and is case-insensitive. Omit to list all creatives.
`page`	integer	Zero-based page index (default: 0).
`size`	integer	Results per page (default: 10, recommended max: 10). Larger windows blow LLM token budgets — only request more when the caller actually needs it.
`contentTypes`	string	Filter by content types (comma-separated). Supported values: IMAGE,VIDEO,GIF,DOCUMENT. Example: IMAGE,VIDEO or IMAGE%2CVIDEO%2CGIF%2CDOCUMENT

Full page, schema and recorded response · machine index

unarchive_ad Writes Unarchive (restore) one or more previously archived ads.

Creates or changes something in the account.

Parameter	Type		Notes
`ids`	array	required	Ad id(s) to unarchive. At least one integer required; pass multiple to restore in bulk.

Full page, schema and recorded response · machine index

update_brand_kit Destructive Update existing brand kit with brand DNA parameters.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`company_url`	string	required	Company website URL or domain (e.g. 'ambient.ai').
`long_description`	string		Company description for marketing use.
`logo_url`	string		Direct URL to a single company logo image (back-compat; prefer logo_urls). Setting a permanent brand logo requires elevated Metadata permissions.
`logo_urls`	array		One or more direct URLs to logo images to set as the brand's permanent logo(s). Each uploaded logo is automatically expanded into clean, persisted variants (transparent, high-res, and both a light and dark version when the mark is a single color) so generated creatives have multiple usable options.
`remove_old_logos`	boolean		When true (default), replace the brand's existing logos with the provided ones; when false, append to them.
`search_keywords`	array		SEO/search keywords for the brand.
`website_background`	string		6-digit hex color (e.g. '#FFFFFF') of the brand website's background. generate_brand_creative uses it as the default dominant background for ads (use_website_background). Set it when the user wants ads to match their site's background or the auto-detected value is wrong.
`colors`	array		Legacy flat color list. Each: {hex (with #), type, name?}.
`fonts`	array		Legacy flat font list. Each: {name, type}.
`primary_colors`	array		Primary color palette.
`accent_colors`	array		Accent color palette.
`extended_palette`	array		Extended grayscale palette.
`semantic_colors`	array		Semantic/status colors.
`primary_font`	string		Primary font family name.
`primary_font_usage`	string		Primary font usage description.
`secondary_font`	string		Secondary/display font family name.
`secondary_font_usage`	string		Secondary font usage description.
`headline_font`	string		Headline font family.
`headline_weight`	string		Headline font weight.
`headline_size`	string		Headline font size range.
`headline_case`	string		Headline text case rule.
`headline_linespace`	string		Headline line spacing.
`headline_letterspace`	string		Headline letter spacing.
`eyebrow_font`	string		Eyebrow font family.
`eyebrow_weight`	string		Eyebrow font weight.
`eyebrow_case`	string		Eyebrow text case rule.
`eyebrow_linespace`	string		Eyebrow line spacing.
`eyebrow_letterspace`	string		Eyebrow letter spacing.
`subtitle_font`	string		Subtitle font family.
`subtitle_weight`	string		Subtitle font weight.
`subtitle_linespace`	string		Subtitle line spacing.
`subtitle_letterspace`	string		Subtitle letter spacing.
`body_font`	string		Body text font family.
`body_weight`	string		Body text font weight.
`body_size`	string		Body text font size.
`body_linespace`	string		Body text line spacing.
`body_letterspace`	string		Body text letter spacing.
`cta_font`	string		CTA/button font family.
`cta_weight`	string		CTA/button font weight.
`cta_size`	string		CTA/button font size.
`cta_linespace`	string		CTA/button line spacing.
`cta_letterspace`	string		CTA/button letter spacing.
`logo_on_dark`	string		Logo color rule on dark backgrounds.
`logo_on_light`	string		Logo color rule on light backgrounds.
`logo_clear_space`	string		Logo clear space rule.
`logo_misuse`	array		List of logo misuse rules.
`button_shape`	string		Button shape description.
`button_background`	string		Button background style.
`button_text_color`	string		Button text color rule.
`button_text_weight`	string		Button text weight.
`button_text_case`	string		Button text case.
`button_padding`	string		Button padding.
`standard_ctas`	array		Standard CTA labels.
`cta_placement`	string		CTA placement guidance.
`tagline`	string		Primary brand tagline.
`tagline_variants`	array		Alternative taglines.
`tagline_placement`	string		Tagline placement guidance.
`visual_motifs`	array		Brand visual motifs.
`promotional_elements`	array		Promotional proof points.
`tone`	string		Brand tone description.
`value_propositions`	array		Core value propositions.
`headline_formulas`	array		Headline writing patterns.
`proof_points`	array		Evidence/social proof items.
`product_names`	object		Product name mapping (e.g. {platform: 'X', vlm: 'Y'}).
`content_hierarchy_order`	array		Content element ordering.
`content_hierarchy_spacing`	string		Section spacing guidance.
`content_hierarchy_text_max_width`	string		Max text block width.
`photography_style`	string		Photography style description.
`photography_subjects`	array		Recommended photo subjects.
`photography_treatment`	string		Photo treatment rules.
`photography_avoid`	string		Photography anti-patterns.
`dos`	array		Brand 'do' rules.
`donts`	array		Brand 'don't' rules.
`primary_layouts`	array		Layout system definitions.
`channel_linkedin`	object		LinkedIn channel config.
`channel_facebook`	object		Facebook channel config.
`channel_display`	object		Display ads channel config.

Full page, schema and recorded response · machine index

update_meta_placements Destructive Set the channel-level Meta placement default of an existing DRAFT wizard campaign's Facebook or Instagram channel: automatic placements (Meta decides where ads serve) or a manua…

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`campaign_id`	integer	required	The unique identifier of the wizard campaign
`channel`	string	required	Which campaign channel to edit. Meta placements only exist on Facebook and Instagram channels. one of: `FACEBOOK`, `INSTAGRAM`
`mode`	string	required	AUTO = automatic placements (Meta decides, clears the manual set). MANUAL = serve only on the placements listed in `placements`. one of: `AUTO`, `MANUAL`
`placements`	array		Placement codes to serve on. Required when mode=MANUAL (at least one); ignored for AUTO. Companion rules apply — see the tool description.

Full page, schema and recorded response · machine index

upload_image_creative Writes Upload images directly to Metadata platform library to create image creatives.

Creates or changes something in the account.

Parameter	Type		Notes
`images`	array	required	Array of image URLs to upload to platform
`names`	array		Optional library name per image, positional against `images`. Pass one whenever you know what the creative is: without it the name is taken from the URL, which for a generated creative is an opaque key like `brand_creative_20260814_211126_66d8a83e.png` and leaves the user unable to tell their creati

Full page, schema and recorded response · machine index

upload_video_creative Writes Upload videos directly to the Metadata platform library to create video creatives.

Creates or changes something in the account.

Parameter	Type		Notes
`videos`	array	required	Array of video URLs to upload to the platform library.
`thumbnail_library_id`	integer		Optional existing imageLibraryId to attach as the video's platform thumbnail.

Full page, schema and recorded response · machine index

add_extensions_to_native_campaign Writes Attach existing library extensions (sitelinks, callouts, etc.) to a native-structure Google Ads / Microsoft Ads campaign, so they actually serve when the campaign launches.

Creates or changes something in the account.

Parameter	Type		Notes
`campaign_id`	integer	required	The wizard campaign id.
`channel`	string	required	Search channel. Ad extensions are a search-channel-only concept. one of: `GOOGLE_ADS`, `MICROSOFT_ADS`
`extension_ids`	array	required	Internal library extension ids from list_extensions. These are the `id` field, NOT `externalId` (the ad platform's own id).

Full page, schema and recorded response · machine index

create_google_business_logo_extension Writes Attach the business logo to a Google Ads text ad for brand recognition.

Creates or changes something in the account.

Parameter	Type
`image_library_id`	integer	required
`business_name`	string
`domain_name`	string

Full page, schema and recorded response · machine index

create_google_callout_extension Writes Add a short non-clickable selling-point snippet (callout) that appears beneath a Google Ads ad — short trust-builder phrases like "Free Shipping", "24/7 Support", "Money-Back Gu…

Creates or changes something in the account.

Parameter	Type		Notes
`callout_text`	string	required	Callout snippet. 1-25 characters.

Full page, schema and recorded response · machine index

create_google_image_extension Writes Attach visual imagery (1-20 product / lifestyle shots) to a Google Ads text ad to lift click-through.

Creates or changes something in the account.

Parameter	Type		Notes
`image_library_ids`	array	required	1-20 image-library IDs.
`alternative_text`	string
`display_text`	string
`destination_url`	string

Full page, schema and recorded response · machine index

create_google_lead_form_extension Writes Add an inline lead-capture form attached to a Google Ads ad — prospects submit their email plus optional contact and work info without leaving the SERP.

Creates or changes something in the account.

Parameter	Type		Notes
`headline`	string	required
`business_name`	string	required
`description`	string
`email_lead_form_user_input_types`	array	required
`contact_lead_form_user_input_types`	array
`work_lead_form_user_input_types`	array
`privacy_policy_url`	string	required
`post_submit_headline`	string	required
`post_submit_description`	string	required
`post_submit_call_to_action_type`	string	required	one of: `VISIT_SITE`, `DOWNLOAD`, `LEARN_MORE`, `SHOP_NOW`
`url`	string		Optional thank-you-page URL.
`call_to_action_type`	string	required	one of: `APPLY_NOW`, `BOOK_NOW`, `CONTACT_US`, `DOWNLOAD`, `GET_INFO`, `GET_OFFER`, `GET_QUOTE`, `GET_STARTED`, `JOIN_NOW`, `LEARN_MORE`, `REGISTER`, `REQUEST_DEMO`
`call_to_action_description`	string	required

Full page, schema and recorded response · machine index

create_google_price_extension Writes Add a price-list shown beneath a Google Ads ad — 3-8 tiles, each with a plan/product name, a short description, a USD amount, and a billing unit.

Creates or changes something in the account.

Parameter	Type		Notes
`price_qualifier`	string	required	Price qualifier shown before the price. one of: `UNSPECIFIED`, `FROM`, `UP_TO`, `AVERAGE`
`price_offerings`	array	required	3-8 price offerings.

Full page, schema and recorded response · machine index

create_google_promotion_extension Writes Add a time-bound promotional offer (sale, discount, promo code) that shows under a Google Ads ad — flag a holiday occasion, pick the discount style (% off or $ off), optionally…

Creates or changes something in the account.

Parameter	Type		Notes
`occasion`	string	required	one of: `NONE`, `BACK_TO_SCHOOL`, `BLACK_FRIDAY`, `BOXING_DAY`, `CARNIVAL`, `CHINESE_NEW_YEAR`, `CHRISTMAS`, `CYBER_MONDAY`, `DIWALI`, `EASTER`, `EID_AL_ADHA`, `EID_AL_FITR`
`url`	string	required
`item`	string	required
`category_type`	string	required	one of: `MONETARY`, `UP_TO_MONETARY`, `PERCENT`, `UP_TO_PERCENT`
`category_amount`	number		USD amount; supply when category_type is MONETARY or UP_TO_MONETARY.
`category_percent`	number		Percent (≤100); supply when category_type is PERCENT or UP_TO_PERCENT.
`details_type`	string	required	one of: `NONE`, `ON_ORDERS_OVER`, `PROMO_CODE`
`details_promo_code`	string		Promo code; supply when details_type is PROMO_CODE.
`details_amount`	number		Threshold USD amount; supply when details_type is ON_ORDERS_OVER.
`start_date`	string	required	ISO-8601 "yyyy-MM-dd'T'HH:mm:ss.SSS'Z'". Future or present.
`end_date`	string	required	ISO-8601 "yyyy-MM-dd'T'HH:mm:ss.SSS'Z'". Future or present.

Full page, schema and recorded response · machine index

create_google_sitelink_extension Writes Add an extra clickable shortcut (sitelink) that appears beneath a Google Ads ad — helpful deep-links like "See Pricing", "Book a Demo", or "Contact Sales" with two short descrip…

Creates or changes something in the account.

Parameter	Type		Notes
`url`	string	required	Destination URL the sitelink points to (non-blank).
`link_text`	string	required	Visible link text. 1-25 characters.
`description1`	string	required	First description line. 1-35 characters.
`description2`	string	required	Second description line. 1-35 characters.

Full page, schema and recorded response · machine index

create_google_structured_snippet_extension Writes Add a labeled list of offerings (structured snippet) shown beneath a Google Ads ad — a category header plus 3-10 short items, e.g.

Creates or changes something in the account.

Parameter	Type		Notes
`header`	string	required	Snippet header. 1-25 characters.
`values`	array	required	3-10 snippet values, each 1-25 characters.

Full page, schema and recorded response · machine index

create_microsoft_business_logo_extension Writes Attach the business logo to a Microsoft Ads (Bing) text ad for brand recognition.

Creates or changes something in the account.

Parameter	Type
`image_library_id`	integer	required
`business_name`	string
`domain_name`	string

Full page, schema and recorded response · machine index

create_microsoft_callout_extension Writes Add a short non-clickable selling-point snippet (callout) that appears beneath a Microsoft Ads (Bing) ad — short trust-builder phrases like "Free Shipping", "24/7 Support", "Mon…

Creates or changes something in the account.

Parameter	Type		Notes
`callout_text`	string	required	Callout snippet. 1-25 characters.

Full page, schema and recorded response · machine index

create_microsoft_image_extension Writes Attach visual imagery (1-20 product / lifestyle shots) to a Microsoft Ads (Bing) text ad to lift click-through.

Creates or changes something in the account.

Parameter	Type		Notes
`image_library_ids`	array	required	1-20 image-library IDs.
`alternative_text`	string
`display_text`	string
`destination_url`	string

Full page, schema and recorded response · machine index

create_microsoft_price_extension Writes Add a price-list shown beneath a Microsoft Ads (Bing) ad — 3-8 tiles, each with a plan/product name, a short description, a USD amount, and a billing unit.

Creates or changes something in the account.

Parameter	Type		Notes
`price_qualifier`	string	required	Price qualifier shown before the price. one of: `UNSPECIFIED`, `FROM`, `UP_TO`, `AVERAGE`
`price_offerings`	array	required	3-8 price offerings.

Full page, schema and recorded response · machine index

create_microsoft_promotion_extension Writes Add a time-bound promotional offer (sale, discount, promo code) that shows under a Microsoft Ads (Bing) ad — flag a holiday occasion, pick the discount style (% off or $ off), o…

Creates or changes something in the account.

Parameter	Type		Notes
`occasion`	string	required	one of: `NONE`, `BACK_TO_SCHOOL`, `BLACK_FRIDAY`, `BOXING_DAY`, `CARNIVAL`, `CHINESE_NEW_YEAR`, `CHRISTMAS`, `CYBER_MONDAY`, `DIWALI`, `EASTER`, `EID_AL_ADHA`, `EID_AL_FITR`
`url`	string	required
`item`	string	required
`category_type`	string	required	one of: `MONETARY`, `UP_TO_MONETARY`, `PERCENT`, `UP_TO_PERCENT`
`category_amount`	number		USD amount; supply when category_type is MONETARY or UP_TO_MONETARY.
`category_percent`	number		Percent (≤100); supply when category_type is PERCENT or UP_TO_PERCENT.
`details_type`	string	required	one of: `NONE`, `ON_ORDERS_OVER`, `PROMO_CODE`
`details_promo_code`	string		Promo code; supply when details_type is PROMO_CODE.
`details_amount`	number		Threshold USD amount; supply when details_type is ON_ORDERS_OVER.
`start_date`	string	required	ISO-8601 "yyyy-MM-dd'T'HH:mm:ss.SSS'Z'". Future or present.
`end_date`	string	required	ISO-8601 "yyyy-MM-dd'T'HH:mm:ss.SSS'Z'". Future or present.

Full page, schema and recorded response · machine index

create_microsoft_sitelink_extension Writes Add an extra clickable shortcut (sitelink) that appears beneath a Microsoft Ads (Bing) ad — helpful deep-links like "See Pricing", "Book a Demo", or "Contact Sales" with two sho…

Creates or changes something in the account.

Parameter	Type		Notes
`url`	string	required	Destination URL the sitelink points to (non-blank).
`link_text`	string	required	Visible link text. 1-25 characters.
`description1`	string	required	First description line. 1-35 characters.
`description2`	string	required	Second description line. 1-35 characters.

Full page, schema and recorded response · machine index

create_microsoft_structured_snippet_extension Writes Add a labeled list of offerings (structured snippet) shown beneath a Microsoft Ads (Bing) ad — a category header plus 3-10 short items, e.g.

Creates or changes something in the account.

Parameter	Type		Notes
`header`	string	required	Snippet header. 1-25 characters.
`values`	array	required	3-10 snippet values, each 1-25 characters.

Full page, schema and recorded response · machine index

list_extensions Read List the account's Google Ads / Microsoft Ads library extensions (sitelinks, callouts, snippets, etc.) with their internal ids.

Returns data. Safe to call freely.

Parameter	Type		Notes
`channel`	string	required	Search channel. Ad extensions are a search-channel-only concept. one of: `GOOGLE_ADS`, `MICROSOFT_ADS`
`types`	array		Optional extension-type filter, e.g. ["SITELINK"].
`name`	string		Optional name filter. Names are auto-derived by the platform, not user-chosen, so prefer filtering by `types`.
`page`	integer		0-based page index. Defaults to 0.
`size`	integer		Page size. Defaults to 50.

Full page, schema and recorded response · machine index

remove_extensions_from_native_campaign Destructive Detach extensions from a native-structure Google Ads / Microsoft Ads campaign.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`campaign_id`	integer	required	The wizard campaign id.
`channel`	string	required	Search channel. Ad extensions are a search-channel-only concept. one of: `GOOGLE_ADS`, `MICROSOFT_ADS`
`extension_ids`	array	required	Internal library extension ids from list_extensions. These are the `id` field, NOT `externalId` (the ad platform's own id).

Full page, schema and recorded response · machine index

add_keywords_to_negative_lists Writes Add keywords to one or more existing Google Ads negative keyword lists.

Creates or changes something in the account.

Parameter	Type		Notes
`negativeListIds`	array	required	IDs of the existing negative keyword lists to extend (from list_negative_keywords_list).
`keywords`	array	required	Keywords to add to every list in negativeListIds. Each item is either a plain string or {text, match_type}.
`match_type`	string		Default match type, used when an item is a plain string or an object that omits match_type. Default PHRASE. one of: `EXACT`, `PHRASE`, `BROAD`
`channel`	string		Ad channel scope. Both search channels are supported. These are PER-CHANNEL resources: GOOGLE_ADS and MICROSOFT_ADS hold separate sets, and one is invisible to the other. Defaults to GOOGLE_ADS when the user does not say which. one of: `GOOGLE_ADS`, `MICROSOFT_ADS`

Full page, schema and recorded response · machine index

add_negative_keyword_lists_to_live_campaigns Writes Attach existing Google Ads negative keyword lists to campaigns that are ALREADY LAUNCHED — a live edit that syncs through to Google.

Creates or changes something in the account.

Parameter	Type		Notes
`campaignIds`	array	required	Wizard campaign ids of the LIVE (already launched) campaigns to update. Pass every campaign the user named in one call. Draft campaigns are skipped with a reason.
`listNames`	array		Names of existing negative keyword lists to attach. Matched exactly (case-insensitive) against the account library; an unmatched name is reported, never substituted.
`listIds`	array		IDs of existing negative keyword lists to attach, as returned by list_negative_keywords_list. Can be combined with listNames.
`mode`	string		'add' (default) merges with the lists already on the campaign. 'replace' makes the campaign's negative lists exactly the ones requested, detaching the rest — only for an explicit user request. one of: `add`, `replace`
`channel`	string		Ad channel scope. Only GOOGLE_ADS is supported — the platform's live-edit endpoint is Google-only. one of: `GOOGLE_ADS`

Full page, schema and recorded response · machine index

create_keywords Writes Create one or more positive keywords in the Metadata platform library.

Creates or changes something in the account.

Parameter	Type		Notes
`keywords`	array	required	Array of keyword strings to create. Example: ["goku", "ssj", "ssj1"]

Full page, schema and recorded response · machine index

create_negative_keywords_list Writes Create a new negative keywords list in the Metadata platform with Google Ads integration.

Creates or changes something in the account.

Parameter	Type		Notes
`name`	string	required	Name of the negative keywords list (string). Example: "Competitor Brands", "Low Intent Keywords"
`keywords`	array	required	Array of keyword strings to block. Example: ["raichu", "blastoise", "charizard"]
`match_type`	string	required	Match type for all keywords. Options: EXACT (exact match), PHRASE (phrase match), BROAD (broad match). Example: 'EXACT' one of: `EXACT`, `PHRASE`, `BROAD`
`channel`	string		Ad channel scope. Both search channels are supported. These are PER-CHANNEL resources: GOOGLE_ADS and MICROSOFT_ADS hold separate sets, and one is invisible to the other. Defaults to GOOGLE_ADS when the user does not say which. one of: `GOOGLE_ADS`, `MICROSOFT_ADS`

Full page, schema and recorded response · machine index

delete_negative_keyword_lists Destructive Permanently delete one or more entire Google Ads negative keyword lists, only when the user has explicitly asked.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`listIds`	array	required	IDs of the negative keyword lists to delete (from list_negative_keywords_list).
`channel`	string		Ad channel scope. Both search channels are supported. These are PER-CHANNEL resources: GOOGLE_ADS and MICROSOFT_ADS hold separate sets, and one is invisible to the other. Defaults to GOOGLE_ADS when the user does not say which. one of: `GOOGLE_ADS`, `MICROSOFT_ADS`

Full page, schema and recorded response · machine index

duplicate_negative_keywords_list Writes Duplicate (copy) a negative keyword list from one search channel to the other: Google Ads to Microsoft Ads (Bing), or Microsoft Ads to Google Ads.

Creates or changes something in the account.

Parameter	Type		Notes
`sourceListId`	integer	required	Id of the negative keyword list to copy, as returned by list_negative_keywords_list.
`sourceChannel`	string		Channel the source list lives on. Default GOOGLE_ADS. one of: `GOOGLE_ADS`, `MICROSOFT_ADS`
`targetChannel`	string		Channel to create the copy on. Must differ from sourceChannel. Default MICROSOFT_ADS. one of: `GOOGLE_ADS`, `MICROSOFT_ADS`
`name`	string		Optional name for the new list. Defaults to "<source name> (<target>)", e.g. "Competitor Brands (Microsoft)".
`dryRun`	boolean		When true, preview the copy (keyword count + BROAD-to-PHRASE conversions) without creating the list.

Full page, schema and recorded response · machine index

experiments_keywords_stats Read Keyword performance data from active experiments with date range filtering.

Returns data. Safe to call freely.

Parameter	Type		Notes
`startDate`	string	required	Start date in ISO 8601 format (YYYY-MM-DDTHH:mm:ss.sssZ). Example: '2025-01-01T00:00:00.000Z'
`endDate`	string	required	End date in ISO 8601 format (YYYY-MM-DDTHH:mm:ss.sssZ). Example: '2025-12-31T23:59:59.999Z'

Full page, schema and recorded response · machine index

get_negative_keywords_list_details Read Fetch the keywords inside a single Google Ads negative keyword list.

Returns data. Safe to call freely.

Parameter	Type		Notes
`listId`	integer	required	ID of the negative keyword list (from list_negative_keywords_list).
`text`	string		Optional substring filter on keyword text.
`page`	integer		0-based page index. Default 0.
`size`	integer		Rows per page. Default 25, max 100.
`sort`	string		Sort criteria in format '<field>,<direction>'. Default 'text,asc'.
`channel`	string		Ad channel scope. Both search channels are supported. These are PER-CHANNEL resources: GOOGLE_ADS and MICROSOFT_ADS hold separate sets, and one is invisible to the other. Defaults to GOOGLE_ADS when the user does not say which. one of: `GOOGLE_ADS`, `MICROSOFT_ADS`

Full page, schema and recorded response · machine index

list_experiment_keywords Read Search and browse keywords running in experiments with simplified response data.

Returns data. Safe to call freely.

Parameter	Type		Notes
`keywordOrExperimentName`	string	required	Search term to filter results by keyword name or experiment name.
`page`	integer		Page number for pagination (0-based). Default is 0.
`size`	integer		Number of items per page. Default is 10.

Full page, schema and recorded response · machine index

list_keywords Read List and search keywords with filtering, sorting, and pagination options.

Returns data. Safe to call freely.

Parameter	Type	Notes
`name`	object	Optional keyword name(s) to search for. Can be a single string or array of strings. Supports partial matching. Omit to list all keywords.
`archived`	boolean	Filter by archived status. Set to false to show active keywords (default), true to include archived keywords.
`page`	integer	Page number for pagination (0-based indexing). Default is 0 for the first page.
`size`	integer	Number of items per page (default: 25, recommended: 25-100).
`sort`	string	Sort criteria in format: field,direction. Options: avgMonthlySearches, lowerPageBid, higherPageBid, name, modifiedDate. Direction: desc (descending) or asc (ascending). Default: modifiedDate,desc one of: `avgMonthlySearches,desc`, `avgMonthlySearches,asc`, `lowerPageBid,desc`, `lowerPageBid,asc`, `higherPageBid,desc`, `higherPageBid,asc`, `name,desc`, `name,asc`, `modifiedDate,desc`, `modifiedDate,asc`

Full page, schema and recorded response · machine index

list_negative_keywords_list Read List and search negative keywords lists by name filter.

Returns data. Safe to call freely.

Parameter	Type		Notes
`name`	object		Optional negative keywords list name(s) to search for. Can be a single string or array of strings. Supports partial matching. Omit to list all negative keywords lists.
`channel`	string		Ad channel scope. Both search channels are supported. These are PER-CHANNEL resources: GOOGLE_ADS and MICROSOFT_ADS hold separate sets, and one is invisible to the other. Defaults to GOOGLE_ADS when the user does not say which. one of: `GOOGLE_ADS`, `MICROSOFT_ADS`

Full page, schema and recorded response · machine index

list_search_terms Read Review the Google Ads search terms report with filtering, sorting, and pagination.

Returns data. Safe to call freely.

Parameter	Type	Notes
`searchTerm`	string	Optional substring filter on the search term text.
`experiment`	string	Optional experiment name substring filter.
`keyword`	string	Optional matched-keyword substring filter.
`matchType`	array	Optional subset of match types to include.
`startDate`	string	Optional ISO-8601 timestamp lower bound (e.g. 2026-06-01T00:00:00.000Z).
`endDate`	string	Optional ISO-8601 timestamp upper bound (e.g. 2026-06-30T23:59:59.999Z).
`page`	integer	0-based page index. Default 0.
`size`	integer	Rows per page. Default 12 (kept small to limit chat-context tokens), max 100.
`sort`	string	Sort criteria in format '<field>,<direction>'. Default 'clicks,desc'. Sortable fields: searchTerm, matchType, experiment, addedExcluded, keyword, impressions, clicks, ctr, averageCpc, spent, conversions, costPerConversion.
`channel`	string	Ad channel scope. Both search channels are supported. These are PER-CHANNEL resources: GOOGLE_ADS and MICROSOFT_ADS hold separate sets, and one is invisible to the other. Defaults to GOOGLE_ADS when the user does not say which. one of: `GOOGLE_ADS`, `MICROSOFT_ADS`

Full page, schema and recorded response · machine index

pause_experiment_keyword Destructive Pause a specific keyword running in an experiment.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`keywordExternalId`	integer	required	The external ID of the keyword to pause.
`wizExperimentId`	integer	required	The experiment ID where the keyword is running.

Full page, schema and recorded response · machine index

remove_keywords_from_campaign Destructive Remove Google Ads keyword groups and/or negative keyword list attachments from a campaign's GOOGLE_ADS channel.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`campaign_id`	integer	required	The unique identifier of the campaign
`channel_id`	integer		Optional fast-path: numeric id of the GOOGLE_ADS channel on this campaign. When provided, the tool skips the full-campaign GET and fetches only this channel's keyword list. Use it when you already have the id from a recent campaign read; omit it and the tool resolves the channel by type.
`channel_keywords`	object		Map of channel type to list of positive keyword group names to remove. Only GOOGLE_ADS is supported. Example: {"GOOGLE_ADS": ["CRM Keywords"]}
`channel_negative_keyword_lists`	object		Map of channel type to list of negative keyword list names to detach from the channel. Only GOOGLE_ADS is supported. Example: {"GOOGLE_ADS": ["Competitor Brands"]}

Full page, schema and recorded response · machine index

remove_keywords_from_negative_list Destructive Remove keywords from a single existing Google Ads negative keyword list, matched by text.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`listId`	integer	required	ID of the negative keyword list (from list_negative_keywords_list).
`keywords`	array	required	Exact keyword texts to remove. Case-sensitive whole-text match.
`channel`	string		Ad channel scope. Both search channels are supported. These are PER-CHANNEL resources: GOOGLE_ADS and MICROSOFT_ADS hold separate sets, and one is invisible to the other. Defaults to GOOGLE_ADS when the user does not say which. one of: `GOOGLE_ADS`, `MICROSOFT_ADS`

Full page, schema and recorded response · machine index

restart_experiment_keyword Writes Restart a specific paused keyword in an experiment.

Creates or changes something in the account.

Parameter	Type		Notes
`keywordExternalId`	integer	required	The external ID of the keyword to restart.
`wizExperimentId`	integer	required	The experiment ID where the keyword is running.

Full page, schema and recorded response · machine index

set_search_channel_final_url_suffix Destructive Set or clear the Final URL Suffix on a wizard campaign's search channel (Google Ads or Microsoft Ads).

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`campaign_id`	integer	required	The unique identifier of the wizard campaign.
`channel`	string	required	Which search channel on the campaign to update. Final URL Suffix is only supported on these two. one of: `GOOGLE_ADS`, `MICROSOFT_ADS`
`final_url_suffix`	string	required	Tracking suffix appended to every final URL on this channel. Format: key=value pairs separated by `&`. Allowed characters: A–Z a–z 0–9 . _ ~ = & { } - or %XX percent-encoded. Pass empty string "" to clear.

Full page, schema and recorded response · machine index

set_search_channel_locations Destructive Set the geographic targeting of a wizard campaign's search channel (Google Ads or Microsoft Ads) — which countries/states the campaign serves in, and which countries to exclude.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`campaign_id`	integer	required	The unique identifier of the wizard campaign.
`channel`	string	required	Which search channel on the campaign to update. Geo targeting is set per channel and only these two carry it. one of: `GOOGLE_ADS`, `MICROSOFT_ADS`
`countries`	array		Country NAMES to target, e.g. ["United States", "Canada"]. REPLACES the channel's current countries. Pass [] to clear (targets all locations). Omit to leave unchanged.
`states`	array		US state NAMES to target, e.g. ["Texas", "California"]. REPLACES the channel's current states. Pass [] to clear. Omit to leave unchanged. US states only.
`excluded_countries`	array		Country NAMES the campaign must NOT serve in, e.g. ["India"]. REPLACES the channel's current exclusions. Pass [] to clear them all. Omit to leave unchanged. Country-level only — states cannot be excluded.

Full page, schema and recorded response · machine index

account_funnel_reports Read Account-level funnel performance across stages.

Returns data. Safe to call freely.

Parameter	Type		Notes
`startDate`	string	required	Start date in ISO 8601 format (YYYY-MM-DDTHH:mm:ss.sssZ). Example: '2025-01-01T00:00:00.000Z'
`endDate`	string	required	End date in ISO 8601 format (YYYY-MM-DDTHH:mm:ss.sssZ). Example: '2025-12-31T23:59:59.999Z'
`reportTab`	string	required	Funnel stage to analyze. Options: REACHED (default), ENGAGED, CONVERTED, MQL, OPPS, WON one of: `REACHED`, `ENGAGED`, `CONVERTED`, `MQL`, `OPPS`, `WON`
`page`	integer		Page number for pagination (0-based)
`size`	integer		Number of results per page. Default: 15, Max: 10000 — raise it to pull the full funnel set in one call.
`sort`	string		Sort criteria. Examples: 'impressions,desc' (default), 'engagement,desc', 'conversions,desc'
`type`	string		Report type. Currently only SOCIAL is supported one of: `SOCIAL`

Full page, schema and recorded response · machine index

account_level_stats Read Account-level statistics and channel breakdown for a timeframe.

Returns data. Safe to call freely.

Parameter	Type	Notes
`timeframe`	string	Timeframe for data aggregation. MONTH=monthly breakdown, QUARTER=quarterly breakdown, WEEK=weekly breakdown, YEAR=yearly breakdown. DO NOT use this parameter when providing startDate and endDate. one of: `QUARTER`, `WEEK`, `YEAR`
`startDate`	string	Start date for custom date range in ISO 8601 format (YYYY-MM-DDTHH:mm:ss.sssZ). Must be used together with endDate. Do NOT provide timeframe when using this. Example: 2025-07-01T00:00:00.000Z
`endDate`	string	End date for custom date range in ISO 8601 format (YYYY-MM-DDTHH:mm:ss.sssZ). Must be used together with startDate. Do NOT provide timeframe when using this. Example: 2025-09-30T00:00:00.000Z

Full page, schema and recorded response · machine index

account_list_performance Read Account list information with engagement metrics and statistics.

Returns data. Safe to call freely.

Parameter	Type	Notes
`page`	integer	Page number for pagination (0-based). Example: page=0 for the first page.
`size`	integer	Number of results per page. Default: 25; raise it (up to 10000) to pull every account list in one call.
`sort`	string	Sorting criteria. Examples: 'triggeredAmount,desc', 'engagedAccounts,desc', 'influencedAmount,desc'. Format: field,direction where direction is 'asc' or 'desc'.
`channels`	string	Comma-separated channel names (e.g., 'LINKEDIN,FACEBOOK,INSTAGRAM,GOOGLE_ADS'). Leave empty for all channels.
`startDate`	string	Start date in ISO 8601 format (YYYY-MM-DDTHH:mm:ss.sssZ). Example: 2025-01-01T00:00:00.000Z. Optional for custom date ranges.
`endDate`	string	End date in ISO 8601 format (YYYY-MM-DDTHH:mm:ss.sssZ). Example: 2025-01-31T23:59:59.999Z. Optional for custom date ranges.

Full page, schema and recorded response · machine index

budget_group_performance Read Performance statistics for a specific budget group by name.

Returns data. Safe to call freely.

Parameter	Type		Notes
`budget_group_name`	string	required	Name of the budget group to analyze (required). The tool will automatically look up the budget group ID. Example: 'Brand Awareness', 'ToFu', 'Lead Gen Q4'
`startDate`	string		Start date for analysis in ISO 8601 format (YYYY-MM-DDTHH:mm:ss.SSSZ, e.g., '2025-01-01T00:00:00.000Z'). Optional - if not provided, no start date filter is applied.
`endDate`	string		End date for analysis in ISO 8601 format (YYYY-MM-DDTHH:mm:ss.SSSZ, e.g., '2025-12-15T23:59:59.999Z'). Optional - if not provided, no end date filter is applied.
`excludeExperimentName`	string		Experiment name to exclude from results. Useful for filtering out specific experiments from the analysis.
`page`	integer		Page number for pagination (0-based). Use for paginating through large result sets.
`size`	integer		Number of experiments per page. Default is 10. Use larger values (up to 10000) for comprehensive analysis — pull the whole set rather than a sample.
`sort`	string		Sort criteria for experiments. Format: field,field,direction. Examples: 'experimentId,name,asc', 'spent,desc', 'cpl,asc'

Full page, schema and recorded response · machine index

deep_funnel_stats Read Comprehensive funnel conversion analysis: funnel conversions, ROI statistics, experiment performance.

Returns data. Safe to call freely.

Parameter	Type		Notes
`funnel_id`	integer	required	The funnel ID to analyze. Example: 35
`startDate`	string		Start date in ISO 8601 format (YYYY-MM-DDTHH:mm:ss.sssZ). Example: '2025-01-01T00:00:00.000Z'
`endDate`	string		End date in ISO 8601 format (YYYY-MM-DDTHH:mm:ss.sssZ). Example: '2025-12-31T23:59:59.999Z'
`opportunityModel`	string		Opportunity attribution model. Options: ALL (default), TRIGGERED, INFLUENCED one of: `ALL`, `TRIGGERED`, `INFLUENCED`

Full page, schema and recorded response · machine index

demographic_country_stats Read Performance metrics grouped by Country.

Returns data. Safe to call freely.

Parameter	Type	Notes
`fromMonth`	string	Start date in ISO 8601 format (YYYY-MM-DDTHH:mm:ss.sssZ). Example: 2025-01-01T00:00:00.000Z. Required for custom date ranges.
`toMonth`	string	End date in ISO 8601 format (YYYY-MM-DDTHH:mm:ss.sssZ). Example: 2025-01-31T23:59:59.999Z. Required for custom date ranges.
`channels`	string	Comma-separated channel names (e.g., 'LINKEDIN,FACEBOOK,INSTAGRAM,GOOGLE_ADS'). Leave empty for all channels.
`page`	integer	Page number for pagination (0-based). Example: page=0 for the first page.
`size`	integer	Number of results per page. Default: 250; raise it (up to 10000) to pull every country in one call.
`sort`	string	Sorting criteria. Examples: 'spend,desc' (default), 'leads,desc', 'cpl,asc', 'impressions,desc'. Format: field,direction where direction is 'asc' or 'desc'.
`groupType`	string	Grouping type - currently only COUNTRY is supported for geographic breakdown. one of: `COUNTRY`

Full page, schema and recorded response · machine index

experiment_performance_stats Read Experiment-level performance statistics from Metadata.io Experiments API.

Returns data. Safe to call freely.

Parameter	Type	Notes
`metrics`	array	List of metrics for secondary ordering when question involves multiple metrics (e.g., ['cpl', 'ctr', 'leads']). Only include when analyzing more than one metric simultaneously. Do NOT include for single metric questions.
`minSpend`	number	Minimum spend threshold. REQUIRED for all campaigns except Brand Awareness. Use for Lead Generation and when campaign type is unspecified.
`minLeads`	integer	Minimum leads threshold. REQUIRED for Lead Generation campaigns. Do NOT use for Brand Awareness campaigns.
`minCpl`	integer	Minimum cost per lead threshold (whole dollars). REQUIRED for Lead Generation campaigns. Do NOT use for Brand Awareness campaigns.
`minAudienceSize`	integer	Minimum audience size to filter experiments
`minClicks`	integer	Minimum clicks threshold. REQUIRED for Brand Awareness campaigns only. Do NOT use for Lead Generation campaigns.
`minCpc`	integer	Minimum cost per click threshold (whole dollars). Do NOT use for Lead Generation or unspecified campaign types.
`size`	integer	Number of results to return. RULES: size=1 for 'top' singular questions, size=requested number for 'top X' questions, size=50 for plural questions (e.g. 'campaigns'), size=30+ for multiple metrics analysis, size=15 as default when not specified. Raise it freely for full-dataset analytics — there is
`sort`	string	Sorting parameter. CRITICAL RULES: 'cpl2Score,desc' for Lead Generation or unspecified campaigns, 'cpc2Score,desc' for Brand Awareness, 'triggeredAmount,desc' for ROI/pipeline questions, 'oppsAmount,desc' for influenced pipeline, 'conversionRate,desc' for conversion rate, 'internalStatus,desc' for a
`goals`	string	Campaign goal type. RULES: 'CPL' for Lead Generation campaigns, 'CTR' for Brand Awareness campaigns. Do NOT include for unspecified campaign types. When multiple metrics involved, use only the first metric's corresponding goal. one of: `CPL`, `CTR`
`name`	object	Experiment name filter, matched as ONE substring per request. Use 'name' field (NOT experimentName) for specific experiment questions. Can be a single string or an array of strings for multiple experiments; an array is queried one name per request and the results are merged. Do NOT add additional pa
`campaignName`	object	Campaign name filter. Use when question is about specific campaign(s). Can be single string or array ['campaign1', 'campaign2'] for multiple campaigns. Do NOT include additional parameters when filtering by specific campaign name.
`audienceName`	string	Filter by target audience name (e.g., 'WTS EMEA Jan25_EMEAPitchbook')
`offerName`	string	Filter by specific offer name (e.g., 'AMER_ZO_EN_HBR Reimagining Work'). Use this field when question mentions specific offers.
`adName`	string	Filter by specific ad name (e.g., 'Sifted_WorkShift_Ad5_Beige'). Use 'adName' field when question asks about specific ad names.
`audienceTypes`	string	Comma-separated list of audience types (e.g., 'Spotlight Retargeting Contacts (Dynamic),Technographic (Aberdeen)')
`imageLibraryName`	string	Filter by creative/image name (e.g., 'JP - Square-Ad15.png'). Use 'imageLibraryName' field when question asks about specific creative names.
`offerTypes`	array	Offer types filter. RULES: ['LP'] for lead gen forms/landing pages questions, ['LP', 'LG'] as default when offer types mentioned. LP=Landing Pages, LG=Lead Gen forms. Should NOT be empty when included.
`channels`	array	Marketing channels to include
`visibilities`	array	Visibility status filter
`launchedExperimentStatuses`	array	Filter experiments by launch status. DEFAULT when omitted: successfully-launched experiments only (Active, WithoutSpend, Paused, Completed) — Failed and Disconnected are EXCLUDED, because a failed/non-launched experiment has 0 leads and is a non-starter, not a 'bottom performer'. Pass values explici
`timeFrame`	string	Timeframe aggregation. CRITICAL RULES: Use ONLY for WEEK or QUARTER questions. NEVER use 'YEAR' - use startDate/endDate instead. If timeFrame is included, DO NOT add startDate or endDate parameters under any circumstances. one of: `WEEK`, `QUARTER`
`startDate`	string	Start date in ISO 8601 format (e.g., '2024-10-18T00:00:00.000Z'). CRITICAL: Do NOT include if timeFrame parameter is used. Use last year of data if period cannot be inferred from question. For specific campaigns/experiments without timeframe, omit this parameter.
`endDate`	string	End date in ISO 8601 format. Must end on last hour of date (e.g., '2024-11-21T23:59:59.999Z'). CRITICAL: Do NOT include if timeFrame parameter is used. For specific campaigns/experiments without timeframe, omit this parameter.

Full page, schema and recorded response · machine index

get_account_opportunities_insights Read Get opportunities for domain(s) with status filtering and pagination.

Returns data. Safe to call freely.

Parameter	Type		Notes
`domains`	array	required	Domain strings. E.g., ['metadata.io']
`status`	array		Filter by status. Omit for all.
`startDate`	string		ISO 8601 format. E.g., '2025-01-01T00:00:00.000Z'
`endDate`	string		ISO 8601 format.
`page`	integer		0-based page number.
`size`	integer		Items per page (default 25, max 10000). Raise it to pull the full set in one call.
`sort`	string		Sort field. Default: lastTouchDate one of: `lastTouchDate`, `status`, `name`, `amount`

Full page, schema and recorded response · machine index

get_account_summary_insights Read Get comprehensive account details for specific domain(s).

Returns data. Safe to call freely.

Parameter	Type		Notes
`domains`	array	required	Domain strings. E.g., ['metadata.io']

Full page, schema and recorded response · machine index

get_converted_leads_summary Read Fast aggregate summary of the account's converted leads (Smart Leads) — counts and rates only, no lead rows.

Returns data. Safe to call freely.

Parameter	Type	Notes
`statuses`	array	Lead statuses to include. The platform Leads tab defaults to ACTIVE. Omit to include all.
`campaignName`	string	Filter by campaign name (partial match).
`experimentName`	string	Filter by experiment name (partial match).
`customAudienceName`	string	Filter by targeted audience name (partial match).
`adName`	string	Filter by the ad (library) name that captured the leads (partial match).
`offerName`	string	Filter by the offer name tied to the leads' experiments.
`jobTitle`	string	Filter by lead job title (partial match).
`email`	string	Filter by lead email (partial match).
`firstName`	string	Filter by lead first name (partial match).
`lastName`	string	Filter by lead last name (partial match).
`businessEmails`	array	Filter by exact business email addresses (OR within the list).
`modifiedBy`	string	Filter by the platform user email who last modified the leads (partial match).
`companies`	array	Filter by exact company names.
`countries`	array	Filter by exact lead country names.
`channels`	array	Filter by acquisition channel(s).
`audienceTypes`	array	Filter by audience type of the leads' experiments (e.g. FIRMOGRAPHIC_INCLUDE, CONTACT_LIST, BUYER_INTENT, NATIVE_LINKEDIN).
`contentTypes`	array	Data-completeness filter: only leads that HAVE this content.
`qualityTypes`	array	Quality-signal filter (opportunities, MQL date/score).
`keyword`	string	Filter by the exact search keyword that captured the leads.
`matchTypes`	array	Filter by keyword match type (search channels).
`startDate`	string	Leads created on/after this date. ISO 8601; date-only is accepted.
`endDate`	string	Leads created on/before this date. ISO 8601; date-only is accepted.
`isFullyEnriched`	boolean	true = only fully enriched leads; false = only not fully enriched.
`isAutoUpdated`	boolean	true = only auto-enriched leads.
`isManuallyUpdate`	boolean	true = only manually enriched/edited leads (platform parameter name is 'isManuallyUpdate').
`syncMA`	boolean	true = only leads synced to marketing automation; false = only unsynced.
`ids`	array	Restrict the summary to specific lead IDs.
`tagIds`	array	Filter by tag IDs on the leads' experiments.

Full page, schema and recorded response · machine index

get_insights_report Read Query account journey insights with filtering, sorting, and pagination.

Returns data. Safe to call freely.

Parameter	Type	Notes
`accountName`	string	Filter by account/company name.
`startDate`	string\|null	ISO 8601 format or null. E.g., '2025-01-01T00:00:00.000Z'
`endDate`	string\|null	ISO 8601 format or null.
`criteria`	object\|null	Criteria object: {"operator":"AND","match":[{field,operator,type,value}]}. Use get_segment_criteria for saved segments.
`page`	integer	0-based page number.
`sort`	string	Sort format: field,direction. Default: lastTouchDate,desc one of: `timeInJourney,desc`, `timeInJourney,asc`, `revenue,desc`, `revenue,asc`, `oppCreatedAmount,desc`, `oppCreatedAmount,asc`, `highIntentDetails,desc`, `highIntentDetails,asc`, `engagementScoreClassificationDelta,desc`, `engagementScoreClassificationDelta,asc`, `engagementScoreClassification,desc`, `engagementScoreClassification,asc`

Full page, schema and recorded response · machine index

performance_metrics Read Granular performance metrics for ads, audiences, creatives, offers, target groups, keywords.

Returns data. Safe to call freely.

Parameter	Type	Notes
`launchedExperimentStatuses`	array	Filter Performance Metrics by experiment launch statuses only.
`page`	integer	Page number to retrieve (0-based indexing)
`size`	integer	Number of results per page. Default 100; raise it freely (up to 10000) for comprehensive full-dataset analysis.
`sort`	string	Sorting criteria (e.g., 'spent,desc', 'cpl,asc', 'experimentId,spent,desc')
`ids`	array	Set of IDs to filter by
`experimentIds`	array	Set of experiment IDs to filter by
`startDate`	string	startDate in ISO 8601 format (YYYY-MM-DDThh:mm:ssZ), e.g., `startDate=2025-06-12T14:30:00.123Z`).
`endDate`	string	endDate in ISO 8601 format (YYYY-MM-DDThh:mm:ssZ), e.g., `startDate=2025-06-12T14:30:00.123Z`).
`oppsStartDate`	string	Start date for opportunities data in ISO 8601 format
`oppsEndDate`	string	End date for opportunities data in ISO 8601 format
`adName`	string	Filter by specific ad name
`keyword`	string	Filter by specific keyword
`campaignName`	string	Filter by specific campaign name
`creativeName`	string	Filter by specific creative name
`customAudienceName`	string	Filter by specific custom audience name
`offerName`	string	Filter by specific offer name
`experimentName`	string	Filter by specific experiment name
`budgetGroupName`	string	Filter by specific budget group name
`channels`	array	Filter by marketing channels
`goals`	array	Filter by campaign goals
`audienceTypes`	array	Filter by audience types
`offerTypes`	array	Filter by offer types. LG=Lead Gen forms, LP=Landing Pages
`visibilities`	array	Filter by visibility status
`adFormats`	array	Filter by ad format types. Applied to ads endpoint automatically.
`minSpend`	number	Minimum spend threshold
`maxSpend`	number	Maximum spend threshold
`minLeads`	integer	Minimum number of leads threshold
`maxLeads`	integer	Maximum number of leads threshold
`minCpl`	number	Minimum cost per lead threshold
`maxCpl`	number	Maximum cost per lead threshold
`minAudienceSize`	integer	Minimum audience size threshold
`maxAudienceSize`	integer	Maximum audience size threshold
`usedInActiveExperiments`	boolean	Filter for ingredients used in active experiments only
`tags`	array	Set of tag IDs to filter by
`accountTagIds`	array	Set of account tag IDs to filter by
`showWithSpend`	boolean	Show only experiments/ingredients with spend data
`experimentList`	string	Experiment list filter
`customFieldIds`	string	Custom field IDs to include
`skipCustomFields`	boolean	Skip custom fields in response
`endpoint`	string	Specific endpoint to query. If not specified, will be auto-determined from query context. Use 'customAudience/group' for target groups. one of: `ads`, `customAudience`, `customAudience/group`, `offer`, `creative`, `keywords`

Full page, schema and recorded response · machine index

query_metadata_analytics_account Read Query the caller's own account CONTEXT from the Metadata.io knowledge base — a rolled-up summary (firmographic, recent 90-day campaign history, open drafts, per-experiment perfo…

Returns data. Safe to call freely.

Parameter	Type		Notes
`question`	string	required	Natural-language question about the caller's account.

Full page, schema and recorded response · machine index

query_metadata_analytics_benchmarks Read Query the Metadata.io marketing knowledge base for cross-account industry benchmarks and priors.

Returns data. Safe to call freely.

Parameter	Type		Notes
`question`	string	required	Natural-language question. Should describe a COHORT (industry × size × channel × format / audience), not a specific company.
`conversation_id`	any		Optional. Pass the value returned from a prior `_benchmarks` call to keep the query context warm. The two knowledge-base tools never share conversation_ids — pass only IDs returned by `_benchmarks` here.

Full page, schema and recorded response · machine index

search_insights_criteria_fields Read Discover available fields for building insights filter criteria.

Returns data. Safe to call freely.

Parameter	Type		Notes
`search_keywords`	array	required	Keywords to search for fields (3-10 recommended). E.g., ['page', 'url', 'visit']

Full page, schema and recorded response · machine index

website_engagement_stats Read Website visitor engagement metrics: session data, page views, visitor behavior analytics.

Returns data. Safe to call freely.

Parameter	Type		Notes
`dateRange`	string	required	Timeframe for engagement data. Options: YESTERDAY, LAST_7_DAYS, LAST_14_DAYS, LAST_30_DAYS, LAST_MONTH, MONTH_TO_DATE (default), LIFETIME. REQUIRED parameter. one of: `YESTERDAY`, `LAST_7_DAYS`, `LAST_14_DAYS`, `LAST_30_DAYS`, `LAST_MONTH`, `MONTH_TO_DATE`, `LIFETIME`
`page`	integer		Page number for pagination (0-based). Example: page=0 for the first page.
`size`	integer		Number of results per page. Default: 15; raise it (up to 10000) to pull the full engagement set in one call.
`sort`	string		Sorting criteria. Examples: 'sessions,desc' (default), 'pageViews,desc', 'uniqueVisitors,desc', 'conversionRate,desc'. Format: field,direction where direction is 'asc' or 'desc'.

Full page, schema and recorded response · machine index

connect_crm Writes Get the OAuth redirect URL to connect a specific CRM system to the Metadata platform.

Creates or changes something in the account.

Parameter	Type		Notes
`crm`	string	required	The CRM system to connect. Must be one of: salesforce, hubspot one of: `salesforce`, `hubspot`
`acknowledge_data_access`	boolean		Set true to confirm the person you are acting for asked to connect this CRM. The response is an OAuth authorization URL that starts a credential grant into their Metadata account. Without this the call is refused.

Full page, schema and recorded response · machine index

create_topup_link Destructive Get a secure payment link the account owner can use to add AI credits.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Takes no arguments.

Full page, schema and recorded response · machine index

get_account_details Read Current user and account information.

Returns data. Safe to call freely.

Takes no arguments.

Full page, schema and recorded response · machine index

get_credit_balance Read Get the current Metadata AI credit balance for the account this token belongs to.

Returns data. Safe to call freely.

Takes no arguments.

Full page, schema and recorded response · machine index

get_credit_transactions Read Get recent AI credit movements for the account this token belongs to, newest first.

Returns data. Safe to call freely.

Parameter	Type		Notes
`page`	integer		Zero-based page number. Defaults to 0 (most recent).
`size`	integer		Rows per page, 1-100. Defaults to 25.

Full page, schema and recorded response · machine index

get_integrations_status Read Get comprehensive integration connection status for all channels, martech platforms, and CRM systems in the Metadata platform.

Returns data. Safe to call freely.

Parameter	Type		Notes
`platformName`	string		Filter by specific platform name (e.g., 'Facebook', 'LinkedIn', 'Salesforce', 'HubSpot')

Full page, schema and recorded response · machine index

get_payment_status Read Check whether a specific credit payment has settled onto the account's ledger.

Returns data. Safe to call freely.

Parameter	Type		Notes
`reference_id`	string	required	The payment's reference, as reported when the purchase was made (e.g. 'maxio:txn:987654'). Found on the transaction row in get_credit_transactions.

Full page, schema and recorded response · machine index

impersonate_account Writes Switch the active Metadata account context to another account so every later tool call operates on THAT account.

Creates or changes something in the account.

Parameter	Type		Notes
`account_id`	string	required	ID of the Metadata account to switch into.

Full page, schema and recorded response · machine index

list_user_accounts Read List accounts the current user has access to, with optional name filtering.

Returns data. Safe to call freely.

Parameter	Type		Notes
`name`	string		Partial account name to search for. Optional for regular users (acts as a filter); required for super users (acts as the search query against the whole platform).

Full page, schema and recorded response · machine index

stop_impersonation Writes Switch back from a client account to your own default Metadata account.

Creates or changes something in the account.

Takes no arguments.

Full page, schema and recorded response · machine index

enrich_company_domain Read Enrich a company DOMAIN with firmographics (industry, employee range, location).

Returns data. Safe to call freely.

Parameter	Type		Notes
`domain`	string	required	Company domain, e.g. 'metadata.io'.

Full page, schema and recorded response · machine index

find_privacy_url Read Locate the privacy-policy URL for a given website domain.

Returns data. Safe to call freely.

Parameter	Type		Notes
`domain`	string	required	Company website URL or domain. Examples: 'metadata.io', 'www.bloomreach.com', 'https://example.com'

Full page, schema and recorded response · machine index

get_current_date Read Return the REAL current date and time (UTC).

Returns data. Safe to call freely.

Takes no arguments.

Full page, schema and recorded response · machine index

resolve_visitor_ip Read Deanonymize a website visitor IP to the COMPANY that owns it.

Returns data. Safe to call freely.

Parameter	Type		Notes
`ip`	string	required	The visitor's IP address (IPv4 or IPv6).

Full page, schema and recorded response · machine index

set_landing_subdomain Destructive Set the subdomain this account's generated landing pages are served from.

Deletes, archives, launches or moves budget. Confirm intent before calling.

Parameter	Type		Notes
`subdomain`	string	required	The subdomain label only, without any zone suffix or scheme ("zoom", not "zoom.ad.live"). Empty string releases the current claim.

Full page, schema and recorded response · machine index

All 160 MCP tools

Audiences and targeting 39

Campaigns and budgets 27

Ads and creative 27

Ad extensions 18

Paid search 18

Analytics and attribution 16

Account and access 10

Utilities 5

Machine-readable