BrightEdge

Use this tool only if the user explicitly requests for 'tracked' in their query. Use AI Brand Presence also know as AI Catalyst (AIB) tools for brand mention and citation data across AI engines (ChatGPT, Perplexity, Google AIO, Google AI Mode). Searches tracked prompts or permanently deletes tracked prompts. Modes: - Search Mode (`prompts`) → Read-only. Returns matching tracked prompts with prompt_id and full prompt text. No confirmation required. - Delete Mode (`prompt_ids`) → Permanently removes tracked prompts and all associated history. Requires confirmation. Use when: - The user wants to stop tracking a prompt but doesn't know the prompt_id → search first, show matches, then delete using the selected prompt_id(s). - The exact prompt_id(s) are already known → delete directly. Do not use when: - Listing tracked prompts or viewing performance → use get_ai_tracked_prompts. - Adding new tracked prompts → use add_ai_tracked_prompts. Important: - Never delete by prompt text alone. Deletion always requires prompt_id(s). - A search may return multiple matching prompts, so let the user choose which prompt_id(s) to remove. Confirmation (Delete Mode): - Requires explicit user confirmation. - Do not set confirmed=true without the user's Yes. - Follow any additional_display_information returned by the tool. Permissions: - Delete requests may be rejected if write access is unavailable or the user has a read-only role. - Relay any tool error exactly as returned and do not imply the deletion succeeded. - Search mode is always allowed. [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

remove_ai_tracked_prompts

Searches tracked prompts or updates an existing tracked prompt's text, intents, and/or topics (free-text labels on that one prompt). Modes: - Search Mode (`search_prompt`) → Read-only. Returns matching tracked prompt(s) with prompt_id and their current prompt text, intents, and topics. No confirmation required. - Update Mode (`prompt_id`) → Updates the given tracked prompt. Requires confirmation. Use when: - The user wants to change a tracked prompt's wording, intent label(s), or topic label(s). - The exact prompt_id is already known → update directly. - The exact prompt_id is not known → search first, show matches, then update using the selected prompt_id. Do not use when: - Listing tracked prompts or viewing performance → use get_ai_tracked_prompts. - Adding new tracked prompts → use add_ai_tracked_prompts. - Removing tracked prompts → use remove_ai_tracked_prompts. - Changing a prompt's country coverage → not supported. Country cannot be edited for an existing tracked prompt. - Renaming/redescribing a prompt group (bucket) itself, or changing its AI engines → use update_ai_tracked_prompt_group. `new_topics` here is a different concept — free-text labels on this one prompt, not the prompt group entity. Important: - Never update by prompt text alone. Updating always requires prompt_id. - `new_intents` and `new_topics` are each the full desired value, not a delta — omit either to leave that field unchanged; pass [] to clear it. - At least one of `new_prompt`, `new_intents`, `new_topics` must be set in Update Mode. - Before updating, the account's prompt usage limit is checked. If this update would exceed the account's configured limit, the tool returns an error instead of updating — tell the user to ask BrightEdge to increase the account's prompt usage limit, or remove some existing tracked prompts, then retry. Confirmation (Update Mode): - Requires explicit user confirmation. - Do not set confirmed=true without the user's Yes. - Follow any additional_display_information returned by the tool. Permissions: - Update requests may be rejected if write access is unavailable or the user has a read-only role. - Relay any tool error exactly as returned and do not imply the update succeeded. - Search mode is always allowed.

update_ai_tracked_prompt

Updates an existing AI Brand Presence prompt group. Use when: - Renaming a prompt group. - Updating its description. - Changing which AI engines (e.g., ChatGPT, Google AIO) evaluate prompts in the group. Do not use when: - Creating or deleting prompt groups → use add_or_delete_ai_tracked_prompt_groups. - Adding or removing prompts → use add_ai_tracked_prompts or remove_ai_tracked_prompts. - Viewing prompt group performance or details → use get_ai_tracked_prompt_groups. Discover / Search: - prompt_group_id defaults to -1 (discover mode) — omit it to list every configured prompt group for this account (plain read). - search_prompt_group_name → looks up prompt groups by (partial) name instead of listing everything (plain read), e.g. 'gift card management'. Just set this and leave prompt_group_id omitted (its -1 default) — search takes priority. - Both skip the update entirely; no confirmation or MCP-Write check applies. Never guess a prompt_group_id — resolve it via one of these first. - engines=[-1] → discover mode for AI engines: returns this account's actually enabled AI engines (read-only) instead of performing the update. Independent of the prompt_group_id / search_prompt_group_name discover checks above — either can be used alone. Validation: - prompt_group_id is checked against this account's actual configured prompt groups before the update is sent, so an unknown or mistyped id returns a clear error instead of an opaque upstream failure. Confirmation: - This is a write operation and requires explicit user confirmation. - Do not set confirmed=true without the user's Yes. - Follow any additional_display_information returned by the tool. Permissions: - The operation may be rejected if write access is unavailable or the user has a read-only role. - Relay tool errors exactly as returned and do not imply the update succeeded.

update_ai_tracked_prompt_group

This is a WRITE tool that adds or removes a tracked COMPETITOR domain for a BrightEdge account, chosen by the `action` parameter (exactly 'add' or 'remove'). Use it whenever the user asks to start or stop tracking a competitor site. It requires user confirmation: call it with confirmed=false first (the default) and it returns a Yes/No confirmation request that makes NO change - relay the options to the user, wait, and only re-call with confirmed=true (same other arguments) once they choose Yes. Never set confirmed=true on your own. A request can be refused (e.g. the account's 'MCP Write' feature bit is off, an add would exceed the account's max tracked-sites limit, or a remove targets a domain that is not tracked); when that happens relay the returned message verbatim and never imply the change succeeded. Adds a competitor domain to a BrightEdge account's tracked competitors, or removes one. `action` selects the operation: - action="add" -> track `domain` as a competitor (optionally with nickname/color/regexp_domain) - action="remove" -> stop tracking `domain` (nickname/color/regexp_domain are ignored). The tool resolves the domain to the tracked competitor internally; if it is not currently tracked, this returns success=false with an explanatory message. WHEN TO USE: - "track amazon.com as a competitor" / "add <domain> to my competitors" -> action="add" - "stop tracking amazon.com" / "remove <domain> from my competitors" -> action="remove" DO NOT USE: - to add a tracked/own domain or keywords -> those are different setup flows KEY PARAMETERS: action - required; exactly 'add' or 'remove'. domain - the competitor domain, e.g. 'amazon.com' (a full URL is accepted and normalized). nickname / color / regexp_domain - add only; color is a plain hex like '#f9a004'. confirmed - leave false on your first call; the tool returns a Yes/No confirmation to show the user. Only re-call with confirmed=true (same other args) after the user chooses Yes. CONFIRMATION: this tool asks the user to confirm before it writes. Your first call returns a confirmation request (status NEEDS_ADDITIONAL_TOOL_CALL) with Yes/No options and makes no change. Present the options, wait, and only on Yes re-call with confirmed=true and the same other args. NOTE: the write can be refused (e.g. the account's 'MCP Write' feature bit is off, or an add would exceed the account's max tracked-sites limit). When refused this returns success=false with a message -- relay it verbatim and do NOT imply it succeeded. [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

add_or_remove_competitor_domain

Use this tool only when the user explicitly asks for engagement analytics, page analytics, site analytics, or page group analytics data. Returns engagement analytics aggregated by page group (site sections or URL patterns). WHEN TO USE: - user asks for analytics by site section, directory, or page group - user wants to compare engagement across grouped pages (e.g. /blog/, /product/) - user asks how a particular section of the site is performing for engagement DO NOT USE: - individual page URL breakdown → use get_page_analytics - site-wide totals → use get_site_analytics - page group GSC click/impression data → use get_google_search_console_pages_performance with query_for="page_groups" - share of voice by page group → use get_share_of_voice_pages - source-level filtering by traffic source (google/bing/etc.) — only available in get_site_analytics [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_page_group_analytics

Use this tool only when the user explicitly asks for engagement analytics, page analytics, site analytics, or page group analytics data. Returns engagement analytics data broken down by individual page URL. WHEN TO USE: - user asks for page-level traffic, visits, or engagement metrics - user wants to see how individual pages are performing in terms of engagement - user asks about bounce rate, time on page, or page views per URL DO NOT USE: - site-wide aggregate metrics → use get_site_analytics - grouped page section performance → use get_page_group_analytics - Google Search Console click/impression data → use get_google_search_console_pages_performance - keyword ranking data → use get_tracked_keywords or get_keywords_for_domain_or_url - source-level filtering by traffic source (google/bing/etc.) — only available in get_site_analytics [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_page_analytics

Use this tool only when the user explicitly asks for engagement analytics, page analytics, site analytics, or page group analytics data. Returns site-wide aggregate engagement analytics for a domain. WHEN TO USE: - user asks for overall site traffic, total visits, or site-level engagement summary - user wants a high-level overview of site performance (sessions, bounce rate, etc.) - user asks how the site is performing overall across all pages DO NOT USE: - page-by-page breakdown → use get_page_analytics - page group breakdown → use get_page_group_analytics - Google Search Console impressions/clicks → use get_google_search_console_overview_summary - keyword tracking data → use get_tracked_keywords_summary [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_site_analytics

Poll an in-flight get_keyword_search job (by job_id) and return its keyword results once ready. Use this ONLY to resume a get_keyword_search call that returned status 'PENDING' with a job_id. Each call polls for a short bounded window; if the job is still running it returns status 'RUNNING' with the same job_id — call again until status is 'Completed'. WHEN TO USE: - get_keyword_search returned status 'PENDING' with a job_id → call this with that job_id - a previous check_keyword_search_status returned status 'RUNNING' → call again to keep polling DO NOT USE: - to start a new keyword search → use get_keyword_search - with a job_id you did not receive from get_keyword_search KEY PARAMETERS: job_id — the job_id returned by get_keyword_search (status 'PENDING'). max_results — capped at 200; the first row is the seed keyword itself.

check_keyword_search_status

Use it for all keyword/domain/url related queries. BrightEdge's datacube is loaded with all the required data. Returns per-keyword rows comparing your domain against up to 4 competitors from BrightEdge DataCube (whole-web index). Use comparison_type to get gap keywords (only competitors rank), shared keywords (both rank), or dominator keywords (competitor outranks you). WHEN TO USE: - "which keywords does my competitor rank for that I don't?" - "show me keywords we both rank for" - "which keywords is competitor outranking me on?" DO NOT USE: - want overlap counts/summary instead of the keyword list → use get_competitor_keyword_summary - single domain keyword list with no competitor → use get_keywords_for_domain_or_url - actual clicks or impressions → use GSC tools KEY PARAMETERS: comparison_type — "only_competitor_ranks", "both_domains_rank", or "competitor_outranks_me" domain — your primary domain (required) competitor_domains — 1–4 competitor domains (required) [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_competitor_keyword_gap_list

This is a WRITE tool that creates a new EVENT (annotation) for a BrightEdge account or deletes existing ones, chosen by the `action` parameter (exactly 'create' or 'delete'). Events are the dated notes shown under Manage Events and overlaid as markers on trend charts - use it when the user wants to record what happened on a date (a migration, redesign, campaign launch) or to remove notes they no longer want. The account is resolved from the logged-in user, so never ask which account to use. It requires user confirmation: call it with confirmed=false first (the default) and it returns a Yes/No confirmation request that makes NO change - relay the options to the user, wait, and only re-call with confirmed=true (same other arguments) once they choose Yes. Never set confirmed=true on your own. Deleting needs event_ids, which you get from get_events - never guess one. A request can be refused (e.g. the account's 'MCP Write' feature bit is off, or the user lacks event-management write access); when that happens relay the returned message verbatim and never imply the change succeeded. `action` selects the operation: - action="create" -> record a new annotation. Needs `description` and `date_occured`; `scope` defaults to 'account' (a site-wide note). Returns the new event_id. - action="delete" -> remove the events listed in `event_ids`. Returns how many were deleted. Get the ids from get_events first; they are not reversible here. WHEN TO USE: - "add an event for the site migration on March 14" / "annotate that we launched the new PDP template yesterday" -> action="create" - "delete that event" / "remove the duplicate annotations" -> action="delete" DO NOT USE: - to read or search events -> use get_events - to change an existing event's text or date - editing is not supported yet KEY PARAMETERS: action - required; exactly 'create' or 'delete'. description / date_occured - create only, both required. Dates accept YYYY-MM-DD. scope / scope_value_id - create only. Leave scope at 'account' unless the user ties the event to a specific keyword, keyword group, competitor domain, page or page group; those need scope_value_id, which you must look up rather than guess. event_ids - delete only; a list of ids from get_events. confirmed - leave false on your first call; the tool returns a Yes/No confirmation to show the user. Only re-call with confirmed=true (same other args) after they choose Yes. CONFIRMATION: this tool asks the user to confirm before it writes. Your first call returns a confirmation request (status NEEDS_ADDITIONAL_TOOL_CALL) with Yes/No options and makes no change. Present the options, wait, and only on Yes re-call with confirmed=true. NOTE: the write can be refused (e.g. the account's 'MCP Write' feature bit is off, or the user lacks event-management write access). When refused this returns success=false with a message -- relay it verbatim and do NOT imply it succeeded.

create_or_delete_event

AEO on-page recommendations for a specific page URL (Answer Engine Optimization). Returns citation counts, top cited competitors, recommended topics, and a detailed comparison for the given page to help improve its AI visibility. WHEN TO USE: - User asks for AEO recommendations for a specific page or URL - User wants to know how to improve a particular page's AI citations - User reports a specific page has N citations vs a competitor and wants to improve it - User asks how to make a specific page more citable in AI responses WHEN NOT TO USE: - User wants an overview across all pages → get_aeo_pages_recommendation_summary - Brand mention leaderboard → get_ai_tracked_prompt_mentions - Citation leaderboard → get_ai_tracked_prompt_citations

get_aeo_page_recommendation

AEO on-page recommendations summary across all pages (Answer Engine Optimization). Returns a paginated list of pages with their citation counts, top competitors, recommended topics, and comparison data to help prioritize AEO improvements. WHEN TO USE: - User wants an overview of AEO recommendations across all pages - User wants to see which pages need the most AEO work - User asks for a summary of on-page optimization opportunities - User wants to convert mentions into citations across multiple pages WHEN NOT TO USE: - User wants recommendations for a specific page → get_aeo_page_recommendation - Brand mention leaderboard → get_ai_tracked_prompt_mentions - Citation leaderboard → get_ai_tracked_prompt_citations

get_aeo_pages_recommendation_summary

Use AI Cube (also known as AI HyperCube) tools for brand mention share, citation share, sentiment, and visibility gap analysis across AI engines (ChatGPT, Perplexity, Google AIO, Google AI Mode). Identify prompts where competitors are cited in AI responses but your brand is not. WHEN TO USE: - When the user would like to see prompts where a specified competitor is cited, but their brand is not mentioned - Find citation gaps where your domain has zero AI citation share - Discover content and link-building opportunities to capture missed AI citations - Prioritize prompts with the highest competitor citation count and no self-brand citation KEY PARAMETERS: - brands: list of brand names to compare (primary brand + competitors) - compare_time_range_start / compare_time_range_end: YYYYMM month range (e.g. 202604, 202605) - cited_domains: list of domains to filter citations by (e.g. ['xyz.com']) - filters: optional additional {field, operator, value} filters - sort_order: desc (default) | asc [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_ai_prompts_visibility_gaps_citations

Use AI Cube (also known as AI HyperCube) tools for brand mention share, citation share, sentiment, and visibility gap analysis across AI engines (ChatGPT, Perplexity, Google AIO, Google AI Mode). Identify prompts where competitors are mentioned in AI responses but your brand is not. WHEN TO USE: - When the user would like to see prompts where their competitors are mentioned but they are not - Find gaps where your brand has zero AI mention share - Discover content opportunities to capture missed AI visibility - Prioritize prompts with the highest competitor mention count and no self-brand mention KEY PARAMETERS: - brands: list of brand names to compare (primary brand + competitors) - compare_time_range_start / compare_time_range_end: YYYYMM month range (e.g. 202604, 202605) - filters: optional additional {field, operator, value} filters - sort_order: desc (default) | asc [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_ai_prompts_visibility_gaps_mentions

Use AI Cube (also known as AI HyperCube) tools for brand mention share, citation share, sentiment, and visibility gap analysis across AI engines (ChatGPT, Perplexity, Google AIO, Google AI Mode). Compare AI citation share across multiple brands for a monthly time window. WHEN TO USE: - Compare citation share for multiple brands (e.g. Brand1 vs Brand2 vs Brand3) over two months - Filter out irrelevant prompt categories (e.g. exclude deal/coupon/sale queries) - Get a ranked list of brands by share_of_citations for a given time period WHEN NOT TO USE: - Single-account weekly citation leaderboard (not comparing multiple brands) → use get_ai_tracked_prompt_citations - Weekly trend/time series of your own citation share → use get_ai_tracked_prompt_citations with trend=true KEY PARAMETERS: - brands: list of brand names to compare (primary brand + competitors) - compare_time_range_start / compare_time_range_end: YYYYMM month range (e.g. 202604, 202605) - filters: optional {field, operator, value} filters to narrow prompt scope - sort_by: share_of_citations (default) | citation_count - sort_order: desc (default) | asc [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_ai_prompt_citations

Use AI Cube (also known as AI HyperCube) tools for brand mention share, citation share, sentiment, and visibility gap analysis across AI engines (ChatGPT, Perplexity, Google AIO, Google AI Mode). Compare AI mention share across multiple brands for a monthly time window. WHEN TO USE: - Compare how often multiple brands are mentioned in AI responses over two months - Get a ranked list of brands by share_of_mentions for a given time period - User asks about "share of mentions", "who is winning", "AI visibility", or "mention percentage" - User asks about AI mention share data availability for a brand or domain (does not require tracked prompts to be configured) KEY PARAMETERS: - brands: list of brand names to compare (primary brand + competitors) - compare_time_range_start / compare_time_range_end: YYYYMM month range (e.g. 202604, 202605) - filters: optional {field, operator, value} filters to narrow prompt scope - sort_by: share_of_mentions (default) | mention_count - sort_order: desc (default) | asc [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_ai_prompt_mentions

Use AI Cube (also known as AI HyperCube) tools for brand mention share, citation share, sentiment, and visibility gap analysis across AI engines (ChatGPT, Perplexity, Google AIO, Google AI Mode). Compare AI brand sentiment (positive/neutral/negative) across multiple brands for a monthly time window. WHEN TO USE: - Compare sentiment distribution across multiple competing brands on specific prompts - Identify which prompts drive positive or negative sentiment for a brand vs competitors - Requires an explicit list of brands to compare WHEN NOT TO USE: - To see how your own brand's sentiment is trending week over week → use get_trended_ai_primary_brand_sentiment (no brands list needed, returns aggregate % per time period) KEY PARAMETERS: - brands: list of brand names to compare (primary brand + competitors) - compare_time_range_start / compare_time_range_end: YYYYMM month range (e.g. 202604, 202605) - filters: optional {field, operator, value} filters to narrow prompt scope [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_ai_prompt_sentiments

Use AI Cube (also known as AI HyperCube) tools for brand mention share, citation share, sentiment, and visibility gap analysis across AI engines (ChatGPT, Perplexity, Google AIO, Google AI Mode). Return the top prompts where your brands (and competitors) are most mentioned in AI responses. WHEN TO USE: - To view the actual AI prompts that are driving my brand's (or competitor's) mentions, ranked by demand - Identify which user queries drive the most AI brand mentions - Prioritize content around high-mention prompts - Understand prompt-level brand visibility in AI engines KEY PARAMETERS: - brands: list of brand names (primary brand and/or competitors) - compare_time_range_start / compare_time_range_end: YYYYMM month range (e.g. 202604, 202605) - filters: optional {field, operator, value} filters to narrow prompt scope - sort_by: be_volume (default) - sort_order: desc (default) | asc [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_ai_prompts

Use this tool only if the user explicitly requests for 'tracked' in their query. Use AI Brand Presence also know as AI Catalyst (AIB) tools for brand mention and citation data across AI engines (ChatGPT, Perplexity, Google AIO, Google AI Mode). AI Brand Presence citation share — domain-level leaderboard (default) or weekly trend. WHEN TO USE: - trend=false (default): domain-level citation comparison / leaderboard across AI engines for a single time window - trend=true: how citation share has changed/trended week over week ("is our citation share going up?", charting citation share over time) - "AIO citation check for [prompt]" — checking citation share for a specific tracked AI prompt (trend=false) - User wants citation share scoped to a specific prompt group / topic (e.g. "Mid-to-low funnel") → pass topic_id_list WHEN NOT TO USE: - Overview tab / brand vs top competitors chart (citations per engine) → use get_ai_tracked_prompt_overview - Mentions → use get_ai_tracked_prompt_mentions - Sentiment → use get_ai_tracked_prompt_sentiments - All prompts (no citation filter) → use get_ai_tracked_prompts - AEO recommendations for a specific page or URL → use get_aeo_page_recommendation - Finding which pages need AEO work to convert mentions into citations → use get_aeo_pages_recommendation_summary KEY PARAMETERS: - trend: false (default) = leaderboard via /get_share_of_citations_data; true = weekly time series via /get_share_of_citations_trend. When trend=true, sort_by/sort_order/pagination_limit/pagination_offset are ignored — the endpoint always returns the full weekly series for the requested range. - brand_category: competitor_and_own (default) | competitor | own | all - sort_by (trend=false only): share_of_citations | citations_count | avg_rank | change - llm: chatgpt | perplexity | google_aio | google_ai_mode | all (default) - prompt_group_id_list: optional list of prompt group (topic) IDs or names to scope the share-of-citations calculation to (e.g. [293360] or ["2026: Retirement"]); omit or pass [] for account-wide results across all prompt groups. Applies to both trend=false and trend=true. - time_range_start / time_range_end: YYYYWW (e.g. 202614); data is weekly — if the user asks for monthly data, multiply the number of months by 4 to get the equivalent number of weeks. Omit both for the latest/most recent available week — do not guess. [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_ai_tracked_prompt_citations

Use this tool only if the user explicitly requests for 'tracked' in their query. Use AI Brand Presence also know as AI Catalyst (AIB) tools for brand mention and citation data across AI engines (ChatGPT, Perplexity, Google AIO, Google AI Mode). Returns prompt group (topic) gains/losses data for a single week (mention, citation, sentiment, rank per group). WHEN TO USE: - User asks about prompt group (topic) performance, gains, or losses at a specific week - Compare prompt groups (topics) by mention/citation/sentiment changes - Filter to specific prompt groups (topics) using topic_id_list WHEN NOT TO USE: - Full tracked prompt list (per-prompt, not per-group) → get_ai_tracked_prompts - Aggregated brand mention share leaderboard → get_ai_tracked_prompt_mentions - Citation leaderboard → get_ai_tracked_prompt_citations - Sentiment trend → get_ai_tracked_prompt_sentiments KEY PARAMETERS: - time_value: single YYYYWW week (e.g. 202625); data is weekly. Omit for the latest/most recent available week — do not guess. - prompt_group_id_list: optional list of prompt group (topic) IDs or names to filter (e.g. [293360] or ["2026: Retirement"]); omit or pass [] for all groups - llm: chatgpt | perplexity | google_aio | google_ai_mode | all (default) - pagination_limit: default is 10; do NOT increase unless the user explicitly asks [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_ai_tracked_prompt_groups

Use this tool only if the user explicitly requests for 'tracked' in their query. Use AI Brand Presence also know as AI Catalyst (AIB) tools for brand mention and citation data across AI engines (ChatGPT, Perplexity, Google AIO, Google AI Mode). AI Brand Presence mention share (brand-level leaderboard). WHEN TO USE: - Aggregated brand-level mention share leaderboard across all prompts (who ranks highest overall) - User explicitly asks for brand mention share or overall mention ranking - User wants mention share scoped to a specific prompt group / topic (e.g. "Mid-to-low funnel") → pass topic_id_list WHEN NOT TO USE: - Overview tab / brand vs top competitors chart (mentions per engine) → use get_ai_tracked_prompt_overview - Per-prompt domain vs competitor comparison → use get_ai_tracked_prompts (sort_by=competitor_brand_mention) - Retrieving/listing tracked prompts → use get_ai_tracked_prompts - Citations → use get_ai_tracked_prompt_citations - Citation trend over time → use get_ai_tracked_prompt_citations with trend=true - Sentiment → use get_ai_tracked_prompt_sentiments - Checking data availability when tracked prompts may not be set up → use get_ai_prompt_mentions - AEO recommendations for a specific page or URL → use get_aeo_page_recommendation - Finding which pages need AEO work to convert mentions into citations → use get_aeo_pages_recommendation_summary KEY PARAMETERS: - brand_category: competitor_and_own (default) | competitor | own | all - sort_by: mention_percentage | count_of_mentions | brand_rank | mention_percentage_change - llm: chatgpt | perplexity | google_aio | google_ai_mode | all (default) - prompt_group_id_list: optional list of prompt group (topic) IDs or names to scope the share-of-mentions calculation to (e.g. [293360] or ["2026: Retirement"]); omit or pass [] for account-wide results across all prompt groups - time_range_start / time_range_end: YYYYWW (e.g. 202614); data is weekly — if the user asks for monthly data, multiply the number of months by 4 to get the equivalent number of weeks. Omit both for the latest/most recent available week — do not guess. [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_ai_tracked_prompt_mentions

Use this tool only if the user explicitly requests for 'tracked' in their query. Use AI Brand Presence also know as AI Catalyst (AIB) tools for brand mention and citation data across AI engines (ChatGPT, Perplexity, Google AIO, Google AI Mode). AI Brand Presence overview — brand vs. top competitors across mentions and citations, per engine. WHEN TO USE: - User asks for the Overview tab / "brand vs top competitors" chart - User wants a single-call snapshot of individual or combined mention and citation share, per AI engine - User wants to compare "all engines" vs. one specific engine (e.g. ChatGPT only) for that comparison WHEN NOT TO USE: - Sortable/filterable/paginated brand mention leaderboard → use get_ai_tracked_prompt_mentions - Sortable/filterable/paginated brand citation leaderboard → use get_ai_tracked_prompt_citations - Sentiment trend over time → use get_ai_tracked_prompt_sentiments with view=trend - Per-prompt drill-down → use get_ai_tracked_prompts - Citation trend over time → use get_ai_tracked_prompt_citations with trend=true KEY PARAMETERS: - llm: chatgpt | perplexity | google_aio | google_ai_mode | all (default). Selecting a single engine scopes every metric in the response (mentions, citations, sentiment) to that engine only. - prompt_group_id_list: optional list of prompt group (topic) IDs or names to scope the overview to (e.g. [293360] or ["2026: Retirement"]); omit or pass [] for account-wide results across all prompt groups - time_range_start / time_range_end: YYYYWW (e.g. 202614); data is weekly — if the user asks for monthly data, multiply the number of months by 4 to get the equivalent number of weeks. Omit both for the latest/most recent available week — do not guess. [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_ai_tracked_prompt_overview

Use this tool only if the user explicitly requests for 'tracked' in their query. Use AI Brand Presence also know as AI Catalyst (AIB) tools for brand mention and citation data across AI engines (ChatGPT, Perplexity, Google AIO, Google AI Mode). AI Brand Presence sentiment trend (over time). WHEN TO USE: - User asks how brand sentiment is trending - User wants sentiment trend scoped to a specific prompt group / topic (e.g. "Mid-to-low funnel") → pass prompt_group_id_list WHEN NOT TO USE: - Overview tab / brand vs top competitors chart (mentions + citations, per engine) → get_ai_tracked_prompt_overview - Mentions → use get_ai_tracked_prompt_mentions - Citations → use get_ai_tracked_prompt_citations - Citation trend over time → use get_ai_tracked_prompt_citations with trend=true - Multi-brand sentiment compare → use compare_ai_brand_sentiment_by_brand (AI Cube) KEY PARAMETERS: - llm: chatgpt | perplexity | google_aio | google_ai_mode | all (default) - prompt_group_id_list: optional list of prompt group (topic) IDs or names to scope the sentiment trend to (e.g. [293360] or ["2026: Retirement"]); omit or pass [] for account-wide results across all prompt groups - time_range_start / time_range_end: YYYYWW (e.g. 202614); data is weekly — if the user asks for monthly data, multiply the number of months by 4 to get the equivalent number of weeks. Omit both for the latest/most recent available week — do not guess. [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_ai_tracked_prompt_sentiments

Use this tool only if the user explicitly requests for 'tracked' in their query. Use AI Brand Presence also know as AI Catalyst (AIB) tools for brand mention and citation data across AI engines (ChatGPT, Perplexity, Google AIO, Google AI Mode). Returns all tracked prompts with full per-prompt data (mentions, citations, sentiment, intent, rank, related prompts). No pre-filtering. WHEN TO USE: - Browse/search all tracked prompts - See full prompt data across mentions + citations + sentiment - Compare your domain vs a competitor domain at the per-prompt level (use sort_by=competitor_brand_mention or self_brand_mention) - User asks to retrieve/list tracked prompts, with or without a competitor angle - User doesn’t ask for a specific gap or leaderboard view - User wants prompts scoped to a specific prompt group (topic) by ID → pass prompt_group_id_list; to filter/search by topic label text instead, use filters with field "topics" WHEN NOT TO USE: - Overview tab / brand vs top competitors chart (mentions + citations, per engine) → get_ai_tracked_prompt_overview - Aggregated brand mention share leaderboard (not per-prompt) → get_ai_tracked_prompt_mentions - Citation leaderboard → get_ai_tracked_prompt_citations - Citation share trend over time (weekly time series, not per-prompt) → get_ai_tracked_prompt_citations with trend=true - Sentiment trend → get_ai_tracked_prompt_sentiments KEY PARAMETERS: - sort_by: competitor_brand_mention | self_brand_mention | self_brand_citation | prompt | intents | self_brand_sentiment - llm: chatgpt | perplexity | google_aio | google_ai_mode | all (default) - prompt_group_id_list: optional list of prompt group (topic) IDs or names to scope results to (e.g. [293360] or ["2026: Retirement"]); omit or pass [] for all groups - filters: field "topics" (operators eq, lk) filters/searches by topic/prompt group label text - time_range_start / time_range_end: YYYYWW (e.g. 202614); data is weekly — if the user asks for monthly data, multiply the number of months by 4 to get the equivalent number of weeks. Omit both for the latest/most recent available week — do not guess. - pagination_limit: default is 10; do NOT increase this unless the user explicitly asks for more results — never call this tool a second time automatically with a higher limit to fetch more [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_ai_tracked_prompts

Use this tool only if the user explicitly requests Share of Voice or SOV data. Returns a summary of Share of Voice (SOV) domains for the current account. WHAT IT DOES: - Summarize tracked domains (your domain + competitors). - Provides summary metrics: SOV score, volume, rank, and time-range comparisons. WHEN TO USE: - To see a summary of all domains in the SOV set. - To analyze tracked competitor domains share of voice results. - To compare a specific competitor domain with the user's domain. - To get a high-level competitive domain summary. DO NOT USE: - For keyword-level SOV → use get_share_of_voice_keywords. - For page-level SOV → use get_share_of_voice_pages. - For keyword gap comparison → use get_competitor_keyword_summary or get_competitor_keyword_gap_list. - For BrightEdge tracked domain rankings → use get_tracked_domains. - For Google Search Console clicks/impressions/CTR → use GSC tools. KEY PARAMETERS: - compare (default "top_domains") → comparison group. - volume_scale (default "brightedge_volume") → scoring scale. - rank_type (default "blended_rank") → ranking type. - keywords_type → include all, branded, or non-branded keywords. - filters → filter by a specific domain. [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_share_of_voice_competitive_domains_summary

Use this tool only if the user explicitly requests Share of Voice or SOV data. Pass a keyword to see all pages ranking for it. Pass a page_url to see all keywords that page ranks for. - If you give a keyword, it shows the pages ranking for that keyword with their SOV scores. - If you give a page_url, it shows the keywords that page ranks for with their SOV scores. WHEN TO USE: - Ask "Which pages rank for keyword X?" and provide keyword=... - Ask "What keywords does this page rank for?" and provide page_url=... - To see competitor pages for a keyword, provide keyword=... and filter by domain. - To see keywords where our page is in the top 3, provide page_url=... and filter by rank less than 4. DO NOT USE: - Do not use this tool for actual Google Search Console clicks. Use GSC tools instead. - Do not use this tool for overall keyword or page SOV summaries. Use keyword/page SOV tools instead. KEY PARAMETERS: - keyword → find pages for that keyword. - page_url → find keywords for that page. - compare (default "top_domains") → defines who to compare against. - volume_scale (default "brightedge_volume") → scoring scale for SOV metrics. - rank_type (default "blended_rank") → blended vs classic ranking. - keywords_type → choose whether to include all keywords or only branded/non-branded keywords. NOTE: If both keyword and page_url are given, the keyword takes priority and the page_url is ignored. [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_share_of_voice_keyword_vs_page_lookup

Use this tool only if the user explicitly requests Share of Voice or SOV data. This tool shows keyword-level SOV metrics. It measures competitive visibility per keyword. WHEN TO USE: - Ask "Which keywords have the highest SOV score?" to see top keywords. - Ask "Show keywords we gained or lost SOV on vs last week" to compare changes. - Ask "How many keywords are in our SOV universe?" and set count=True. DO NOT USE: - Do not use this tool for actual clicks, impressions, or CTR. Use get_google_search_console_keyword_performance. - Do not use this tool for tracked keyword rankings. Use get_tracked_keywords. - Do not use this tool for untracked keyword research. Use get_keywords_for_domain_or_url. - Do not use this tool for "keywords ranking for [keyword phrase or URL]" — use get_keywords_for_domain_or_url. - Do not use this tool for page-level SOV. Use get_share_of_voice_pages. - Do not use this tool for keyword↔page drill-down. Use get_share_of_voice_keyword_vs_page_lookup. - Do not use this tool for SERP feature presence, unlinked features, missed SERP connection opportunities, SERP feature gains/losses, or SERP visibility movement. Use get_domain_or_url_serp_features. KEY PARAMETERS: - time_range_start / time_range_end → set both to get comparison fields. - domain_to_filter → convenience shorthand for filtering by a single domain. [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_share_of_voice_keywords

Use this tool only if the user explicitly requests Share of Voice or SOV data. This tool shows page-level SOV metrics. It measures competitive visibility per landing page. It does NOT show clicks or raw rank. WHEN TO USE: - Ask "Which pages have the highest SOV score?" to see top pages. - Ask "Show pages that gained or lost the most SOV this month vs last month" to compare changes. - Ask "How many unique pages rank in our SOV universe?" and set count=True. - To see competitor pages, filter by domain name. - To see only our own pages, filter by domain_type=1. DO NOT USE: - Do not use this tool for actual page clicks, impressions, or CTR. Use get_google_search_console_pages_performance. - Do not use this tool for site section or directory gains/losses. Use get_pages_gaining_or_losing_trends. - Do not use this tool for estimated visits per page. Use get_tracked_keyword_estimation. - Do not use this tool for keyword-level SOV. Use get_share_of_voice_keywords. - Do not use this tool for keyword↔page drill-down. Use get_share_of_voice_keyword_vs_page_lookup. KEY PARAMETERS: - compare (default "top_domains") → defines which competitors to compare against (own domain + tracked competitors, or just own domain). - volume_scale (default "brightedge_volume") → scoring scale for SOV metrics (Google volume-weighted or BrightEdge volume-weighted). - rank_type (default "blended_rank") → blended vs classic ranking. - keywords_type → choose whether to include all keywords or only branded/non-branded keywords in the SOV calculation. [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_share_of_voice_pages

Use it for all keyword/domain/url related queries. BrightEdge's datacube is loaded with all the required data. Returns SERP feature presence and opportunities for a domain across the full web index — covers featured snippets, AI Overviews, People Also Ask, and more. Covers any domain, not limited to tracked keywords. WHEN TO USE: - "which keywords trigger a featured snippet / AI Overview / People Also Ask?" - "show me SERP feature opportunities for my domain" - "which SERP features am I winning or losing?" - "which SERP features are unlinked to my pages?" - "where are we appearing in SERP features but not properly linked / missed connection opportunities" → set serp_features_view="unlinked_features", is_branded=False - "which SERP features did we gain or lose / full visibility movement in SERP features" → set serp_features_view="gains_and_losses" - "keywords that gained or lost AI Overview coverage between two periods" → set serp_features_view="gains_and_losses" - "AI Overview keyword opportunities" or "find keywords where AI Overview exists but we're not featured" → set serp_features_view="win_opportunities" DO NOT USE: - user says "my tracked keywords" and wants per-keyword SERP rows → use get_tracked_keyword_serp - user says "my tracked keywords" and wants aggregate SERP feature counts → use get_tracked_keyword_serp_summary - just need keyword rankings without SERP feature data → use get_keywords_for_domain_or_url - actual GSC clicks → use GSC tools - "AIO citation check for [prompt]" where the subject is a tracked AI prompt, not a URL → use get_ai_tracked_prompt_citations KEY PARAMETERS: serp_features_view — "my_rankings" (default), "all_features", "win_opportunities", "unlinked_features", or "gains_and_losses" [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_domain_or_url_serp_features

This tool returns BrightEdge SEO recommendations analysis data. Return an aggregate summary of SEO Recommendations across all pages for a domain. WHEN TO USE: - "give me a recommendations summary for my domain" - "how many recommendations do I have?" - "what are the top issues across my site?" - "show me a breakdown of recommendations by category" - "which pages need the most attention?" DO NOT USE: - single-page recommendations → use get_seo_page_recommendations - keyword/traffic analytics → use DCX or GSC tools [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_seo_pages_recommendation_summary

This tool returns BrightEdge SEO recommendations analysis data. Return a paginated list of keywords with their SEO recommendation data. WHEN TO USE: - "show me keyword recommendations across my domain" - "which keywords have the most recommendations?" - "what SEO issues exist at the keyword level?" - "show me keywords sorted by search volume with recommendation counts" DO NOT USE: - single-keyword detail → use get_seo_keyword_recommendations - page-level recommendations → use get_seo_pages_recommendation_summary - keyword rank/traffic data → use get_tracked_keywords or GSC tools [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_seo_keywords_recommendation_summary

NOTE: DataCube "Research Keywords" tool — discovers NEW keyword ideas topically related to a seed keyword, covering the full web index (not limited to a domain's rankings or your tracked list). Given a seed keyword, returns a list of topically related keywords with BrightEdge search volume, month-over-month and year-over-year volume change, CPC, competition, keyword difficulty, and word count. The first row is the seed keyword itself. WHEN TO USE: - "research keywords related to [seed]" / "find keyword ideas for [seed]" - "what keywords are topically related to [seed]?" - "give me long-tail / related keywords for [seed] with search volume and trend" - discovering content opportunities around a topic before it is tied to any domain DO NOT USE: - keywords a specific domain or URL already ranks for → use get_keywords_for_domain_or_url - your BrightEdge tracked keyword list → use get_tracked_keywords - semantically related keywords tied to your account's own rankings / page context → use get_related_keywords (SearchIQ) KEY PARAMETERS: keyword — a single seed keyword. Only one keyword is supported (the value is not split on commas). max_results — capped at 200 (the UI limit); the first row is the seed keyword itself.

get_keyword_search

Use it for all keyword/domain/url related queries. BrightEdge's datacube is loaded with all the required data. NOTE: DataCube tool — covers the full web index. For keywords in your BrightEdge tracked list, use get_tracked_keywords instead. Returns keyword rankings and movement data for a domain or URL from BrightEdge DataCube. Covers all keyword types: tracked, non-tracked, branded, non-branded. WHEN TO USE: - "what keywords does my site rank for?" / "keywords ranking for [keyword phrase or URL]" - "show me keywords gaining/losing rank/traffic/volume" → set keyword_change_metric - "top keywords by traffic/volume for this domain or URL" - "which non-branded keywords gained the most rank on our product pages" → set keyword_change_metric="rank", is_branded=False, url=<product page URL> DO NOT USE: - "my tracked keywords" or "my BrightEdge list" → use get_tracked_keywords - actual clicks, impressions, real traffic → use get_google_search_console_keyword_performance - SERP feature presence → use get_domain_or_url_serp_features - page/folder gains or losses → use get_pages_gaining_or_losing_trends KEY PARAMETERS: keyword_change_metric — set to "rank", "traffic", or "volume" to see which keywords moved on that metric between start_date and end_date. Omit for current snapshot rankings. order — "desc" returns the biggest gainers first (default); use "asc" when the user asks for losers, drops, or declines. [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_keywords_for_domain_or_url

Use it for all keyword/domain/url related queries. BrightEdge's datacube is loaded with all the required data. Returns keyword overlap counts between your domain and up to 4 competitors (e.g. "how many keywords do I share with competitor X?"). Covers the full web index — any domain, tracked or not. WHEN TO USE: - "how many keywords do I share with a competitor?" - "give me a competitive keyword coverage summary" - "keyword count comparison across domains" DO NOT USE: - want the actual keyword list, not just counts → use get_competitor_keyword_gap_list - avg rank or page-1/2/3 counts for tracked domains → use get_tracked_domains - share-of-voice / market share score → use SOV tools KEY PARAMETERS: domain — your primary domain (required) competitor_domains — 1–4 competitor domains (required) [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_competitor_keyword_summary

Use it for all keyword/domain/url related queries. BrightEdge's datacube is loaded with all the required data. This tool shows which pages are gaining or losing visibility. It is based on rank-based estimated traffic from BrightEdge DataCube. It does NOT show actual Google clicks. WHEN TO USE: - "which sections of my site are gaining/losing visibility?" - "which folders are declining in rankings?" - "show me page gains or page losses by directory" DO NOT USE: - actual click/impression data per page → use get_google_search_console_pages_performance - per-keyword rank movement rows → use get_keywords_for_domain_or_url with keyword_change_metric="rank" - competitive market-share loss by page → use get_share_of_voice_pages KEY PARAMETERS: trend — "gaining" or "losing"; infer from user prompt ("lost/declining/dropped" → "losing", "gained/growing/improving" → "gaining"). For both trends, make two separate calls. [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_pages_gaining_or_losing_trends

This tool returns BrightEdge SEO recommendations analysis data. Return SEO Recommendations for a specific keyword, including competitor comparison. WHEN TO USE: - "what recommendations do I have for this keyword?" - "how can I improve my ranking for this keyword?" - "show me SEO issues for a specific keyword" - "who are the top 10 ranking pages for this keyword?" - "how does my page compare onpage to competitors for this keyword?" DO NOT USE: - domain-wide keyword summary → use get_seo_keywords_recommendation_summary - page-level recommendations → use get_seo_page_recommendations - keyword rank history → use get_tracked_keyword_history REQUIRED PARAM: - keyword [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_seo_keyword_recommendations

This tool returns BrightEdge SEO recommendations analysis data. Return SEO Recommendations for a specific page URL. WHEN TO USE: - "what recommendations do I have for this page?" - "how can I improve this URL?" - "show me SEO issues for this page" - "what should I fix on this page?" DO NOT USE: - domain-wide recommendation summary → use get_seo_pages_recommendation_summary - keyword rank data → use get_keywords_for_domain_or_url - page traffic data → use get_google_search_console_pages_performance [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_seo_page_recommendations

Use this tool only when the user explicitly asks for Google Search Console (GSC) data. Returns keyword-level click/impression/CTR/position data from Google Search Console with period comparison. WHEN TO USE: - user asks for keywords/queries/search terms as the primary subject, without anchoring to one specific URL - keyword groups or keyword counts from GSC - user explicitly asks to break a named page down by individual query — must use exact phrasing like "list the queries" or "keyword-level breakdown" (use query_for="keyword_groups"), not just "top search queries for this page" DO NOT USE: - request opens with a specific page/URL and asks about its performance → use get_google_search_console_pages_performance first - tracked keyword rankings from BrightEdge → use get_tracked_keywords - keyword research / DataCube universe → use get_keywords_for_domain_or_url - keywords driving a page by SOV score → use get_share_of_voice_keyword_vs_page_lookup - for a single aggregate summary (total clicks, impressions, CTR, position across everything) → use get_google_search_console_overview_summary - SERP feature coverage changes between periods (e.g. "keywords that lost AI Overview coverage") → use get_domain_or_url_serp_features KEY PARAMETERS: query_for — "keywords" or "keyword_groups" (keywords driving a specific page) return_counts_only — True for totals only, no rows (only applies when query_for="keywords") [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_google_search_console_keyword_performance

Use this tool only when the user explicitly asks for Google Search Console (GSC) data. Returns page-level click/impression/CTR/position data from Google Search Console with period comparison. WHEN TO USE: - actual clicks, impressions, CTR, or position per page/URL from GSC - page groups or page counts from GSC - ANY request that names a specific page/URL first, even if it also mentions "queries" or "search terms" — start here for page-scoped requests and return page-level totals unless the user explicitly asks to break down by individual keyword/query - "how is this page performing", "top pages", "page performance for X" DO NOT USE: - user explicitly asks for the LIST of individual keywords/queries (not just page totals) → use get_google_search_console_keyword_performance - page SOV score or competitive visibility → use get_share_of_voice_pages - pages ranking for a keyword by SOV score → use get_share_of_voice_keyword_vs_page_lookup - site directory organic gains/losses → use get_pages_gaining_or_losing_trends KEY PARAMETERS: query_for — "pages" (default), "page_groups". return_counts_only — True for totals only, numeric output with no rows. [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_google_search_console_pages_performance

Use this tool only when the user explicitly asks for Google Search Console (GSC) data. Returns a GSC click/impression/CTR/position summary. Not broken down by keyword or page. KEY PARAMETERS: - query_for="overview_summary" → preset windows (week, month, quarter); no date range needed (default) - query_for="time_range_based_summary" → time-range-based comparison; requires compare_time_range_start and compare_time_range_end WHEN TO USE: - user wants a GSC snapshot without specifying explicit date ranges - user names explicit date windows (e.g. "this month vs last month", "Q1 vs Q2") - headline metrics comparison for any period DO NOT USE: - estimated traffic or revenue → use get_tracked_keyword_estimation_summary - tracked keyword ranking KPI summary → use get_tracked_keywords_summary - keyword or page grids → use get_google_search_console_keyword_performance or get_google_search_console_pages_performance [IMPORTANT] ALWAYS WARN USERS: If the response contains an actions array, always inform the user using the label value and follow the action instruction.

get_google_search_console_overview_summary