Searchable

Create a user-authored prompt topic (the same row the in-app Topics tab creates). A name the project already uses (ignoring case) is refused with `conflict` and the existing topic's id (`existingTopicId`) — use that topic instead. Requires the `prompt:draft` capability — topics organize prompts, the same surface add_prompts / update_prompt gate on. Pass `dryRun: true` to preview it first.

create_topic

DEPRECATED — call `get_brand_profile` with `include=domain_authority` instead (same data). This name is scheduled for removal; see the Searchable API changelog at https://docs.searchable.com.

get_domain_authority

Google Analytics 4 website traffic. `report=ai_referrals` (per-LLM-host visitor series from ChatGPT/Copilot/Gemini/Perplexity/… — use for AI-referral attribution), `trend` (daily users/sessions/pageviews), `top_pages` (top landing pages by sessions), or `sources` (channel-group + source breakdown). Returns `ga4_not_connected` if GA4 isn't linked.

get_ga4_traffic

Google Search Console organic search performance. `report`: `overview` (default — clicks/impressions/CTR/avg position + prior-period comparison), `top_queries`, `top_pages`, `trend` (daily series + totals), `countries`, `devices`, `opportunities` (quick-win queries: position 4-15, >=100 impressions, <5% CTR), `position_buckets` (top 3 / page 1 / page 2 / beyond 20). Window: `period` (7d|28d|30d|90d|16m, default 30d) or explicit `startDate`+`endDate` (YYYY-MM-DD); GSC data lags ~3 days. `limit` applies to top_queries/top_pages/countries/opportunities. Returns `gsc_not_connected` if no GSC site is linked.

get_gsc_performance

Generate + publish a shareable report and return its public URL. `reportType`=sentiment|visibility|combined (default sentiment). Works on a read-only connection, as sharing does in the app; the workspace's sharing policy and a plan with report sharing still apply. Without `confirm:true` returns a dry-run preview (no report created).

generate_report

Get a single article/content piece with full content — HTML/markdown body, outline, FAQs, images, schema markup, internal/external links, and all metadata. Use to pull content for uploading to a site.

get_article

Read a project's live Associated Sources registry or scouting suggestions. Registry isEffective reports whether each live row exactly matches the latest daily published attribution rule; suggestions include durable refresh state.

get_associated_sources

Get the current UTC date/time plus pre-computed window starts (7/30/90/365 days). Call this BEFORE interpreting relative time like 'last month' or picking a days window — never compute dates yourself.

get_current_date

The Prompt Universe brief for a project: which buyer question types matter at each funnel moment for a business model, which are decisive for it, and how many prompts each moment and kind should carry. Planning input only — it calls no model and creates nothing. Write the prompts yourself, then add them with import_prompts (pass dryRun: true first to check them against the prompt gates).

get_prompt_brief

A prompt's wording history, newest first, alongside its current revision number. Read-only — unaffected by the pitch freeze, which only blocks mutations.

get_prompt_revisions

AI visibility for a project. `group_by` selects the view: `summary` (default — overall visibility score), `platform` (per-engine breakdown), `prompt` (per-prompt, paginated), `location` (per country/city), `date` (per-report time series for trend/uplift/before-after analysis; default 90d, branded/unbranded filters only). Per-topic visibility lives in `get_topic_analysis` — the only topic view. `unbranded`/`branded` restrict by prompt brand state; `topicId`/`country`/`locationId` narrow every view except `date`. Responses echo `appliedFilters` (`{}` when unfiltered); `summary`/`platform` also return `availableFilters` (the project's topics + countries). An interactive Visibility Snapshot card also renders for `summary`/`platform` calls in Apps-capable clients (score + trend + platform bars, brand/topic/location controls in fullscreen, filtered calls labeled); it stays hidden on a `platform`-filtered call and on `prompt`/`location`/`date`, where its numbers would misrepresent the answer.

get_visibility

Add prompts you wrote from get_prompt_brief to a project as tracked prompts, as the app adds prompts: they're checked from the project's next scheduled run and count toward the prompt allowance. Starts no Prompt Universe run and calls no model. Every candidate is held to the Prompt Universe code gates BEFORE anything is written — no em/en dash, no semicolon, the market's spelling, no service request at the owning moment, no near-duplicate of each other or of the project's existing prompts — and one gate refusal refuses the whole batch, naming each refused candidate and why; the whole batch is then checked against the allowance, and one that does not fit adds nothing and returns `quota_exceeded`, so tell the user they're out of room and archive some prompts (`archive_prompt`) or import fewer. The writes themselves are NOT atomic: each prompt commits on its own and re-checks the allowance, so a batch can still stop partway if another write takes the room. Every response carries `outcomes`, one row per candidate (created / failed / not_attempted) — on a partial failure, retry only the rows that are not `created`. Up to 10 per call. Requires the `prompt:draft` capability. Pass `dryRun: true` to run the gates and the allowance check without adding anything.

import_prompts

Answer a WHOLE diagnostic question in one call — runs the multi-tool chain server-side and returns the measured signals together. Use this FIRST for 'why did X change' / 'where do we stand' questions instead of chaining the granular tools yourself. `question=visibility_drop` (score trend + per-platform current-vs-previous + share-of-voice shift + weakest topics — for any 'why did visibility fall/change' question), `competitive_position` (share-of-voice standing with ranks and deltas + the topics where a rival leads), `citation_gap` (brand citation trend + the domains AI answers actually cite + per-platform citation share). Returns MEASUREMENTS ONLY plus an `interpretation` list of what the numbers cannot settle — it never names a cause. A signal whose upstream failed appears in `unavailable` with a reason, never as a zero. Drill further with the individual tools once you know where to look.

investigate

List articles / content pieces for a project (merges content-v2 + legacy). Shows title, status, type, word count, and source. Filter by `status` (draft|published|editing|all). Use get_article to pull one piece's full content.

list_articles

List the Searchable projects this connection can access — name, domain, ID, and `dataAvailableFrom` (the earliest date with visibility data, null when the project has none yet). Call this FIRST: every other tool needs a projectId, and the horizon tells you how wide a window is worth requesting. All data is project-scoped.

list_projects

Create, update, or delete an Associated Source; track/dismiss a scouting suggestion; or queue a paid suggestion refresh with action=refresh_suggestions. Shared social/community platforms are always URL-scoped even when domain scope is requested; returned source.scope is authoritative. Requires write scope. Without confirm:true returns a dry run. Never retry refresh automatically: a later retry is another paid scout requiring fresh confirmation. Poll get_associated_sources with view=suggestions and only report completion when latestRefreshCompleted=true. Live registry changes reach analytics only after the daily attribution publication.

manage_associated_source

Read the prioritized fix/write/audit opportunities Searchable surfaced for a project (derived from visibility, sentiment, sources, traffic, prompts, site health, articles). Read-only. Defaults to active; pass `status` / `includeResolved` to widen, `source` / `impact` to narrow. Returned in product priority order (status → impact → newest) with global stats. Each row's `id` is what `update_opportunity_status` takes.

get_opportunities

Per-URL rollup of the three series for one page: AI-answer activity citing it (inlineCitations = answer-text links, sourceUses = consulted as a source, with daily sourceUses buckets), AI-referral sessions landing on it, and AI-crawler hits fetching it — with platform/bot breakdowns. `url` is a full URL or a bare path like `/pricing` (resolved against the project domain). Traffic series read `available: false` when LLM Analytics isn't connected, or reason `external_url` for a host outside the project's verified domains; the citation series degrades the same way on a lookup failure. One call instead of three for 'how is this page doing in AI'.

get_page_metrics

Raw, per-response AI-answer data for one prompt — platform, response date, the full response text, whether the brand was mentioned, competitors mentioned, and the source URLs the engine consulted (sources[] — pages used for generation, not inline citations), one entry per AI response. Not an aggregate — use get_visibility or get_share_of_voice for scores/rankings. The default concise format lists a short snippet per answer; response_format=detailed prints every answer in full, so use it to read or quote what an engine actually said.

get_prompt_answers

The project's prompt catalog as configured — the monitored (tracked) prompts with their type, intent, branded flag, topics, tracked locations, and keyword volume/difficulty; paused (untracked) prompts; and the AI-suggested queue awaiting review (status="suggested" — proposals from research runs and the agent that nobody has accepted or rejected yet, with the reason each was proposed). Configuration data, not performance — use get_visibility group_by=prompt for per-prompt visibility scores, and get_prompt_answers for raw AI responses. The summary always reports all three counts so you can see pending suggestions from any call. An interactive Prompts card (a compact prompt table inline, the full metrics table with topic/branded filters and sortable columns in fullscreen) ALSO renders beside your reply in Apps-capable clients.

list_prompts

Query fanout — the sub-queries AI models generate when answering tracked prompts. Without `prompt_id`: the project aggregate (most frequent fanout queries, per-prompt counts, per-model stats; paginated). With `prompt_id`: every deduped sub-query for that one prompt with per-query frequency + query type. Paid feature — returns `plan_upgrade_required` if not entitled.

get_query_fanout

Reactivate an archived project, as the app's Reactivate action does: its tracked prompts are checked again from the next scheduled run. A pitch is refused with `pitch_not_supported`. Requires a connection minted with the `project:manage` capability (not just `write`). Pass `dryRun: true` to preview it first.

reactivate_project

Read a full Searchable documentation page by slug (from search_docs), optionally narrowed to one section heading.

read_doc

Re-sync the project's configured sitemap — re-discovers it and reconciles monitored pages (add/update/remove) in the background; returns the job id. Read the result with get_site_health (view=pages). Requires the `write` scope and write access to the project (a workspace viewer is refused), and is blocked for pitch projects. Without `confirm:true` returns a dry-run preview (no refresh started).

refresh_sitemap

Restore an archived prompt, as the app's Restore action does, so it is tracked again and counts toward scheduled AI-visibility collection. Prompts added with add_prompts or import_prompts are tracked already and need no restore. Requires the `prompt:activate` capability. Pass `dryRun: true` to preview it first.

restore_prompt

Restore a prompt's wording to an earlier revision by appending a NEW revision with that text (history is never rewritten). Only wording changes — tracking, topics and locations are untouched. `expectedRevision` is required. Requires the `prompt:draft` capability. Pass `dryRun: true` to preview it first.

restore_prompt_wording

Top cited source domains from AI answers — rank, citation counts, usage %, models, top content types. Paginated + sortable. Use to find which external sites AI models cite most; drill in with get_source_detail.

search_sources

Search the Searchable product documentation (metric definitions, setup guides, integrations, API reference, troubleshooting). Returns matching pages with slugs — follow up with read_doc. Use for any 'how does Searchable …' or 'what does this metric mean' question.

search_docs

DEPRECATED — call `get_sentiment` with `view=history` instead (same data). This name is scheduled for removal; see the Searchable API changelog at https://docs.searchable.com.

get_sentiment_history

DEPRECATED — call `get_sentiment` with `view=competitors` instead (same data). This name is scheduled for removal; see the Searchable API changelog at https://docs.searchable.com.

get_sentiment_competitors

Brand vs competitor share of voice — mention-share percentage, rank among all tracked entities, and day-over-day (today vs yesterday) point change. `group_by=date` returns the daily share series for the brand plus its top competitors instead (use it for 'is our share growing', trend, and before/after questions). Mentions/citations counts and the full comparison table aren't part of this view (mirrors the in-app Share of Voice sheet) — use get_competitors for both. `unbranded`/`branded` restrict by prompt brand state; `topicId`/`country`/`locationId` narrow. Responses echo `appliedFilters` (`{}` when unfiltered) plus `availableFilters` (the project's topics + countries). An interactive Share of Voice card also renders in Apps-capable clients (your share as the headline, its movement, and share bars for the rest, with brand/topic/location controls in fullscreen; filtered calls labeled); it stays hidden on `group_by=date`, whose daily series it can't draw.

get_share_of_voice

Products surfaced in AI shopping/product-carousel answers. `view=summary` (default): ranked product list (rank, mentions, dominant platform) + activation-rate summary. `view=timeseries`: daily activation-rate/product-count trend. Pass `productId` (a summary row's `id` — the product title) for one product's detail — occurrences, vendor/pricing rows, per-platform breakdown, and up to 50 recent responses (`productId` overrides `view`). Summary accepts `unbranded`/`branded` + `topicId`/`country`/`locationId` narrowing and echoes `appliedFilters` (`{}` when unfiltered) plus `availableFilters`. `available: false` means no shopping data yet — a data-presence state, not an error; results appear once tracked prompts trigger an AI shopping carousel. An interactive Shopping Shelf card also renders on the summary view (image, title, price, rating, platform; topic/brand/location controls in fullscreen) in Apps-capable clients.

get_shopping_visibility

Technical + AEO site health. `view=pages` (default): the monitored-page inventory, ONE server-paginated page at a time (`limit`/`offset`, `search` by URL/slug/title, `sortBy`; audited pages first) — latest technical/AEO scores, open-issue count, last-audited date, `pagination.total`, and header `summary` counts unless `includeSummary:false`. `view=audits`: per-page scores for audited pages + averages. `view=issues`: issues grouped by severity (`status`=open|resolved|all). `view=issue_details`: issues grouped by type with fix guidance (`issueType`=technical|content). An interactive Site Audit card also renders: pages/audits draw the paginated inventory (server-side search + Prev/Next paging, a per-page Audit button calling trigger_audit, a Sync-website button calling refresh_sitemap); issues views draw findings ranked by severity, expandable to fix guidance.

get_site_health

Drill into cited sources. `level=domain` (+ `domain`): domain summary, or `include=urls` (cited URLs on it) / `include=responses` (AI responses citing it). `level=url`: with `url` → single-page citation analytics, or `include=content` → cached page markdown + entity mentions; without `url` → all cited URLs across the project (optional `contentType`).

get_source_detail

How the project's top source domains trend over time — a per-domain time-series of the external sites most used as sources in AI answers, over the window (default 30 days). Renders an interactive Sources card: the trend line inline, expanding to the top source URLs. Just call it with the projectId — only pass `days` for a non-default window (7 / 90 / 365).

get_source_trends

Per-topic AI visibility. `view=topics` (default) gives each topic's brand mention rate, competitive rank, contributing prompt count, and the AI-identified attribute strengths and gaps behind the number; `view=prompts` drills into ONE topic (`topicId` required) for the per-prompt brand rank and which competitors outrank it — the 'which questions are we losing' view; `view=heatmap` gives average brand position per topic x AI platform. The ONLY per-topic surface — use it for any topic/subject-area/category question, including where the brand wins or loses and what to fix. `unbranded`/`branded` restrict by prompt brand state. An interactive Topic Analysis card also renders for `view=topics` (a visibility bar per topic, expandable to strengths, gaps, and top rival) in Apps-capable clients; it stays hidden on the other views, whose shapes it can't draw.

get_topic_analysis

Start a technical + AEO site audit for the project's pages (all tracked pages by default, or the given `pageIds`); returns the job id. Results land on get_site_health (view=audits). Requires the `write` scope and write access to the project (a workspace viewer is refused), consumes the monthly audit quota, and is blocked for pitch projects. Pass `dryRun: true` to preview it first (no audit started).

trigger_audit

Update the project's knowledge-base brand fields (name, headline, positioning, description, category, influence, language, websiteUrl, caseSensitive, brandNameDisambiguation, topics); each field says what it is in the app, and only the fields you pass change. Does NOT touch connectedEntities, eeatData, or the project's default location. Requires the `knowledge:manage` capability. Pass `dryRun: true` to preview it first.

update_brand_profile

Edit a tracked competitor as the app's competitor sheet does: rename it, change its website domain, make it a direct or a SERP competitor, and add or remove the other names it goes by (name variations), which is how its mentions are recognised in AI answers. Changing the domain, or making it direct, re-matches its past mentions in the background. The result lists its name variations, so a dryRun with no fields shows them. A variation that is another competitor's name is refused with `conflict` (merge the two in the app instead), and removing a name it does not have is refused with `invalid_argument`; either way nothing is changed. Take the competitor id from get_competitors. Requires the `competitor:match` capability and a manage-tier role, as in the app. Pass `dryRun: true` to see the change without making it.

update_competitor

Mark an opportunity completed or dismissed (optionally with a reason/feedback). Opportunities are not a pitch-workspace feature: an ACTIVE pitch project refuses with `pitch_not_supported`, matching get_opportunities. Requires the `opportunity:manage` capability. Pass `dryRun: true` to preview it first.

update_opportunity_status