Bloomiro

Get Mcp Guide

Use only when the request is ambiguous, a Bloomiro field needs explanation, or you do not know which intent-based tool applies. Do not call before an obvious request. Returns a compact routing answer and relevant resource URIs, not the full documentation. Bloomiro is an intelligence and action layer for measurable visibility experiments, not an oracle.

Get Project Context

Use when the site is unknown, the user has multiple sites, a domain must be resolved to its site_id, or integrations and permissions must be checked. Do not call when a previous result already supplied the correct site_id. Accessible sites are paginated with a default limit of 20.

Get Search Performance

Use for Google Search Console queries, pages, clicks, impressions, CTR, positions, and traffic-change investigation. Choose one dimension unless both are explicitly needed. Defaults to 10 rows and no daily trend. Increase limit up to 50, add filters, or set include_trend only for deliberate deeper analysis.

Get Web Analytics

Use for connected Google Analytics (GA4) traffic: sessions, users, engagement, acquisition, content, events, conversions, countries, and tracking setup. Default view=overview returns totals plus three compact highlights. Use one focused view for depth. Optional filters include or exclude country, device, channel, source, medium, or page, with all filters combined. Not Search Console: sessions are not Google clicks.

Get Site Insights

Use to read saved site-wide crawl findings or the combined visibility summary. Returns Search Readiness (seo_score) and AI Readiness (geo_score) as secondary metrics; lead with top_findings and pages needing attention. Saved scans default to overview_only=true. Set overview_only=false for findings, include_pages=true only for deliberate page detail, and limit to bound returned arrays.

Get Page Insights

Use to list saved page analyses or read one saved analysis. Returns issues first; Search Readiness (seo_score) and AI Readiness (geo_score) are secondary. Lists and issue arrays default to 10 items. Full page content is an on-demand resource and is never embedded in the normal response.

Search Sitemap

Use proactively when the user asks which pages exist, whether a topic or path is already on the site, or what the sitemap declares. On the connected project, pass query (slug, title, or topic) to search live sitemap paths, stored crawl pages, and every connected CMS, labeled by connection. Fuzzy-matches exact paths, close slugs, and titles. If a matching live path exists, recommend updating it; do not create a duplicate. CMS page lists are not the full domain. For a competitor or public host, pa

Get Robots Txt

Use for crawlability, robots directives, sitemap references, or search and AI crawler access. Returns a compact policy summary by default. Set user_agent for a specific crawler and include_raw only when exact bounded directives are needed. Robots rules do not guarantee indexing, rankings, AI mentions, or citations.

Get Domain Rating

Use to retrieve the current Ahrefs Domain Rating for a domain or URL. Do not treat it as a traffic, ranking, or AI visibility score. Always preserve the returned Ahrefs attribution and license URL when reporting it. This live lookup costs zero Bloomiro credits.

Get Page Speed

Aliases: lighthouse test, PageSpeed Insights, PSI, test page speed, run Lighthouse, Core Web Vitals, CWV, site speed, LCP/INP/CLS, performance score. Live Google PageSpeed Insights / Lighthouse on a URL or the site homepage. Default view=summary returns compact scores, field/lab CWV, and top opportunities with audit ids. Request view=opportunities, diagnostics, metrics, or view=audit with one audit_id only when needed. Never dumps full Lighthouse JSON. Never saved in Bloomiro; each call re-runs

Get Ai Presence

Use to retrieve measured AI visibility: visibility_score, mention_rate, brand_citation_share, sentiment_score, avg_mention_position, entity_metrics (brand vs competitors), citations, and per-prompt results. These are measured outcomes on tracked prompts, not predictions of future mentions. view=overview and view=history return every prompt group plus a compact all-time history series (dates, visibility_score, mention_rate, brand citations) and charts[] (line/bar/pie). Do not page groups. view=ru

Get Ai Answer

Focused alias for get_ai_presence view=result. Use with result_id to retrieve one tracked model answer, mentions, competitors, citations, and compact AI Search Trace metadata without loading a full run.

Get Ai Search Trace

Use when the user asks what a tracked AI model such as ChatGPT searched, which underlying queries it used, or what search angles were observed. Pass result_id for one result, or provider/run_id/prompt_group_id to read a compact batch. Returns only provider-observed metadata from Postgres, never raw response streams or private reasoning. Exact queries are in items[].search_trace.queries. Do not substitute Google AI Overview/SERP search ideas from get_ai_search_opportunities.

Get Community Sources

Use to understand where AI systems already get information: Wikipedia, Reddit, Quora, YouTube, Facebook, LinkedIn, review/directory (G2/Capterra) pages, and related community sources that answers cite. Helps diagnose off-site visibility gaps; onsite edits alone may not move answers shaped by these sources. view=overview returns compact totals, per-platform hit/miss, and by_model counts (no static platform essays). view=urls lists pages with citation_count, models, and prompt_ids (default 10, nex

Get Competitor Insights

Use to read tracked competitors, one saved scan, visibility gaps, or content coverage. Lead with content_coverage and visibility_gap; treat seo_score/geo_score as technical readiness only. Scan detail defaults to overview_only=true and returned arrays default to 10 items. Increase limit or request pages only when deeper evidence is needed.

Find Opportunities

Use after or alongside evidence reads to surface the highest-priority experiment candidates across GSC, crawl, page, AI presence, and competitor data. This is the broad opportunity engine, not the fan-out ideas tool. For Google AI Overview/SERP fan-out ideas, call get_ai_search_opportunities. For exact provider-observed model searches, call get_ai_search_trace. Returns saved or computed opportunities with stable references. Treat items as hypotheses to prioritize by impact, confidence, effort, a

Get Work Items

Read the site's Action Center markdown board: sections, short notes, and tracked checklist items. Check in early when planning, continuing, or closing a loop (skip pure factual lookups); assume the human may have edited the board. Prefer optimizing existing work over inventing parallel tasks. Tracked work is checklist items (text, optional note, priority, visibility, item_id), evidence-backed; see bloomiro://reasoning/action-center. Linked to the tasks table. Not a billing plan. Hard limit 30000

Run Analysis

Start site, page, visibility, AI presence, competitor, or cited content pattern analysis. cited_content_patterns runs in the background, costs 1 credit on successful completion, analyzes up to 20 exact URLs cited at least twice, and replaces the prior result. Read it later with get_ai_search_opportunities view=content_patterns. Ask before paid work. Do not sleep, wait, or poll in the same conversation.

Manage Prompts

Create or edit AI presence prompt groups and buyer prompts. One call: create_group with name and optional prompts[]. Or create_group then add_prompt with prompt_group_id. List ids with get_ai_presence view=prompts. Archive/restore many with archive_prompt or restore_prompt plus prompt_ids (max 50). remove_group deletes a group. Auto-run: set_auto_run / pause_auto_run (slot 14:00|02:00 Europe/Stockholm; cadence daily|twice_weekly|weekly|fortnightly; days[] prefer ISO 1=Mon…7=Sun).

Manage Page Scans

Delete saved page scans by analysis_id or analysis_ids (max 50). List ids with get_page_insights view=list.

Manage Competitors

Add, update, archive, or restore tracked competitors. Read/scan via get_competitor_insights or run_analysis.

Manage Work Items

Optimize Action Center: update/toggle/move items, delete_items with item_ids (max 50), replace_block for notes; append only new non-duplicates. Prefer chat for one-off advice. Hard limit 30000 chars.

Request Gsc Indexing

Use only to ask Google to crawl a URL that belongs to the connected Search Console property. This reaches Google and does not guarantee indexing or ranking. Do not use it to inspect index status or search performance.