Fodda Knowledge Graphs MCP Server

Surfaces strategic deflection and narrative tension by comparing Wall Street analyst questions against executive responses across quarterly earnings calls. Use when investigating corporate transparency, executive dodge/deflection patterns, or uncovering unaddressed operational risks (e.g., 'Where are retail CEOs deflecting analyst questions on margin pressure?', 'Show analyst-management divergence for hospitality brands'). Sibling routing: Use get_company_earnings for standard single-ticker quarterly records and get_earnings_intelligence for multi-company strategic initiatives; use get_earnings_divergence specifically to analyze tension between analyst scrutiny and executive deflection. Returns clustered divergence themes, verbatim question/response pairs, company deflection counts, and deflection severity metrics. OUTPUT REQUIREMENT: For multi-company thematic findings, present results in a structured evidence table citing: | Company (Ticker) | Reporting Quarter | Speaker (Analyst & Firm / Exec) | Verbatim Quote / Finding |. Clearly distinguish Fodda's synthesized theme labels (e.g. "Pilot Trap") from what analysts or executives actually said on the call.

Cross-company thematic earnings intelligence from the knowledge graph and web sources. Use for multi-company comparisons ("what are hotel companies saying about labor costs?"), industry-level queries, sector filters, or strategic corporate activity categories (activity: "sustainability" | "marketing" | "retail" | "technology"). For corporate ESG, circular economy, and climate initiatives (e.g. "What sustainability commitments did apparel companies make this quarter?"), pass activity="sustainability" or rely on server-side intent routing to avoid conflation with analyst margin sustainability questions. Multi-company and sector queries return active initiatives along with confirmed_absent brands (checked tickers with confirmed silence / no initiatives discussed). For single-brand earnings, brand_tracker includes earnings automatically. For per-ticker structured analysis (analyst concerns, activity breakdown, validated consumer trends), use get_company_earnings instead — it reads the canonical truth layer. Results may include "knowledge_graph" or "web_supplemental" provenance. OUTPUT REQUIREMENT: For multi-company thematic findings, present results in a structured evidence table citing: | Company (Ticker) | Reporting Quarter | Speaker (Analyst & Firm / Exec) | Verbatim Quote / Finding |. Clearly distinguish Fodda's synthesized theme labels (e.g. "Pilot Trap") from what analysts or executives actually said on the call.

Check creator revenue share and consultation earnings balance for an authenticated Human Agent on Fodda. Use when an expert creator asks about their earnings, consultation fees, or royalty payouts (e.g., 'How much have I earned on Fodda?', 'Check my expert payout balance'). Requires an authenticated expert account. Sibling routing: This tool is strictly for human creator monetization—do NOT use for public company SEC filings or corporate earnings (use get_company_earnings for per-ticker records or get_earnings_intelligence for multi-company comparisons). Returns creator financial status including lifetime_earnings_usd, pending_payout_usd, completed_consultations, and payout_status.

Initiate background research on the expert's public work and domain insights to support expertise and voice modeling. Research runs asynchronously in the background for 1-2 minutes. Do NOT wait now; proceed immediately to the Tone of Voice study in chat.

Use when seeking candidate AI-powered Human Agent (Digital Twin) expert models or Synthetic Analysts across specialized domains or research questions (e.g. 'Who has expertise in luxury retail?' or 'What would consumer culture analysts say about this?'). Searches Fodda's catalog of 4,000+ domain specialists to return 2–3 ranked expert models with their analytical focus, methodology, credibility anchors, and expert ID for consultation via consult_human_agent or consult_analyst.

[Deprecated: Legacy alias of get_specialist_intelligence. Prefer search_graph for topic-routed multi-graph searches, or consult_human_agent / consult_analyst for direct strategist consultation.] Parallel search across specialist strategist knowledge graphs curated by domain strategists and boutique studios (culture, youth trends, commerce, media). Sibling routing: For consumer domain trends use get_domain_intelligence; for standalone statistics use search_statistics; for direct 1-on-1 expert advisory use consult_human_agent. Returns structured specialist frameworks, strategic commentary, and curated trend evidence.

Step 1b in Human Agent onboarding: registers expert profile details (name, role, knowledge area, consultation call rate, optional bio/headshot) and records explicit acceptance of Terms of Service and Privacy Policy. Prerequisite: The expert must review and explicitly accept Fodda Terms (https://www.fodda.ai/terms) and Privacy Policy (https://www.fodda.ai/privacy) with termsAccepted: true before progress can be saved. Workflow sequence: Advances to expert_onboarding_research for standard onboarding or submit_mcp_source for BYO-MCP. Returns basic_info_saved status and next step instructions.

Step 3 in standard Human Agent onboarding: submits the expert's structured voice study and domain expertise map derived from conversational analysis and public writing. Prerequisite: Must follow submit_basic_info and expert_onboarding_research. Workflow sequence: Submitting this analysis prepares the thematic taxonomy; immediately follow this call with get_detected_themes to review extracted themes. Sibling routing: This is for standard onboarding—BYO-MCP onboarding skips this step and uses submit_mcp_source instead. Returns submission confirmation and advances onboarding status to theme review.

Step 2 in the Bring-Your-Own-MCP (BYO-MCP) expert onboarding flow: connects to and probes an external public HTTPS MCP endpoint to evaluate live tools and infer domain expertise for a Human Agent digital twin. Prerequisite: submit_basic_info must be called first. Alternative: If the expert does not have an external MCP endpoint, use the standard knowledge-graph path via expert_onboarding_research instead. Side effects: Sends an external live probe to the target MCP server to discover tools and capabilities; does not finalize onboarding until finalize_byo_mcp_onboarding is invoked. Returns discovered tools count, tool names, derived expertise topics, and recommended next step.

Use when seeking role-specific strategic guidance, executive decision frameworks, or historical perspectives from synthetic domain analysts, corporate C-Suite executive personas, or classic historical thinkers grounded in specialist knowledge graphs (e.g., 'Ask the Brand CMO persona how to evaluate retail media ROI', 'Consult John Ruskin on craftsmanship in luxury fashion'). Supports multi-turn dialogue via session_id, corporate brand binding via company, and comprehensive deep research (deep: true). Sibling routing: Use consult_human_agent for authorized Human Agents (Human Agents) (e.g. Ben Dietz); use consult_analyst for synthetic personas, C-Suite roles, and classic thinkers; use list_analysts to browse available personas. Returns synthesized analyst response, cited graph insights, and recommended next moves.

Use when consulting an authorized Human Agent (Digital Twin) — an AI expert model created directly with the named practitioner's consent, participation, and curated knowledge graph. The AI expert model answers in the practitioner's voice and analytical methodology for one-off questions or multi-turn engagements (pass session_id back to continue). Call find_expert or list_analysts first to find the right expert ID and distinguish active Human Agents (consultable immediately via consult_human_agent) from On-Request Human Agents (prospective advisors where introductions can be requested via request_expert_intro). Sibling routing: Use consult_human_agent for interactive AI consultation with active Human Agents; use consult_analyst for synthetic/classic personas. Returns practitioner-grounded analysis, coverage status (in/adjacent/out), source attribution, and structured next moves.

Long-form thought leadership studio: transforms research findings into an in-depth 800–1,200 word LinkedIn article grounded in Fodda knowledge graphs and corporate disclosures. Use when commissioning publication-ready editorial essays, thought leadership articles, or detailed strategic perspectives (e.g., 'Write an 800-word LinkedIn article on retail agentic commerce with data receipts'). Performs an evidence sweep across 3–5 sub-themes, quantitative statistics, and analyst thesis evaluation. Sibling routing: Use draft_linkedin_post for concise, high-impact social posts (150–300 words); use draft_linkedin_article for structured, multi-section long-form articles. Returns an evidence pack, strategic outline, and explicit composition contract for the host assistant to draft.

Draft a LinkedIn post about any topic, grounded in Fodda's expert knowledge graphs. Use when the user says "draft a LinkedIn post about…", "write a post on…", "turn this into a LinkedIn post", or wants social content backed by receipts. Returns a curated EVIDENCE PACK (claims with named companies, typed sources, and real URLs — never constructed) plus a strict composition contract; YOU write the post from it. Every claim is verifiable, thin coverage is flagged honestly, and dropped themes are logged with reasons. Bills as one content call; identical re-requests within 24h serve from cache free. (Trend evidence extraction + tone/formatting prompt execution + draft compilation.)

Translate cultural trends, brand challenges, and consumer insights into a 2026 Strategy-to-Activation Comms & Retail Media Plan. Generates an agency-grade media deliverable with T1-T5 tasks, 3-layer shopper audience cards (with mandatory exclusions), Amazon/Walmart digital shelf intelligence, tactical placement tables with [LIVE] provenance tags, and a machine-readable activation manifest.

Create a presentation-ready data visualization from research findings. Available chart types: "cultural_shifts" (From→To transitions), "competitive_compass" (brands on 2 axes), "trend_constellation" (network of related trends), "implication_ladder" (Signal→Trend→So What→Do What), "innovation_pathway" (Now→Near-Term→Future), "opportunity_map" (2×2 white space analysis). Returns a branded SVG that renders directly in the chat. (Data payload transformation + SVG chart rendering.)

Use when commissioning an executive document from an analyst or Human Agent—such as a strategic briefing, market memo, competitive assessment, or publication-ready article—rather than conversational chat (e.g., 'Have Ben Dietz write a strategic memo on retail media networks'). Workflow routing: Discover available analysts and their specific commissionable offerings first via list_analysts or consult them interactively via consult_human_agent / consult_analyst. Execution & follow-up: Requests may process asynchronously; returns a job_id that can be polled to completion using check_deliverable_status. Returns job confirmation, estimated completion time, and deliverable status.

Use when a user wants to request a facilitated advisory introduction or 1-on-1 consultation with an On-Request Human Agent or active specialist across 4,150+ indexed topics, or when an expert does not maintain a direct public calendar link (e.g., 'I want to speak with Peter Abraham about bicycle retail strategy'). Sibling routing: Use consult_human_agent for immediate AI-driven consultation with active Human Agents; use find_expert to discover specialists; use request_expert_intro when the user seeks a human-to-human advisory introduction. Returns immediate request confirmation, concierge tracking reference, and expected email follow-up timeline.

Check if market data gathering is complete and retrieve the results. Call this after get_supplemental_context — poll every 5-10 seconds until status is COMPLETE or FAILED. (Async supplemental query status poll.)

Federated empirical data layer: queries 110+ live institutional and government feeds (FRED, BLS, US Census, World Bank, OECD, WTO, IMF PortWatch, etc.) for real-time macroeconomic indicators, demographic shifts, global trade flows, container shipping & port throughput, maritime chokepoints (Suez, Panama), and retail spending. When citing US retail sales, cite official Headline Retail & Food Services (NAICS 44X72 / FRED RSAFS, ~\$738B) as the primary retail figure; clearly distinguish from Retail Trade Excl. Food Services (NAICS 44000 / FRED RSXFS, ~\$646B). When citing consumer spending, report live Monthly Personal Consumption Expenditures (BEA Table T20805 / FRED PCE, ~\$22.3T through August) with MoM and YoY growth rates rather than quarterly aggregates. Use when verifying macroeconomic benchmarks, checking consumer spending metrics, grounding supply chain or trend hypotheses with empirical statistics, or when knowledge graph coverage is thin. Sibling routing: Use search_graph or get_domain_intelligence for curated consumer trends and brand case studies; use get_supplemental_context when you need authoritative government statistics, port/logistics telemetry, or regional/country economic context. Returns a structured 5-bucket evidence bundle (demand_signals, economic_context, market_data, financial_reporting, demographic_context) with full source provenance.

Use when verifying a factual claim, hypothesis, or strategic assertion against a named practitioner's curated knowledge graph via their AI-powered Human Agent (Digital Twin) model. Evaluates whether the practitioner's documented research and domain evidence supports or refutes the claim. If analyst_id is omitted, automatically routes to the most relevant active Human Agent covering the domain. Sibling routing: Use verify_claim to evaluate a claim against a specific practitioner's knowledge graph and receive a structured verdict; use verify_market_claim for multi-graph algorithmic 5-gate verification across 100+ graphs and earnings. Returns a structured verification object with verdict ("confirms", "contradicts", "partial", or "no_coverage"), confidence ("full", "partial", or "thin"), one_line summary, rationale, cited sources, and expert attribution.

Use when verifying a factual market claim against 100+ knowledge graphs, primary research citations, and corporate earnings divergence through 5 sequential kill-gates (grounding, contradiction, freshness, corroboration, divergence). Sibling routing: Use verify_market_claim for objective, multi-source empirical data verification and contradiction scoring; use verify_claim to evaluate a claim against a specific practitioner's curated knowledge graph via their AI-powered Human Agent model. Returns structured verdict ([VERIFIED], [REFUTED_BY_EVIDENCE], [INSUFFICIENT_EVIDENCE], [UNGROUNDED_SPECULATION]), supporting evidence citations, and contradiction scores.

Check the current user's account status: API call balance, plan, enabled/disabled graphs, and profile info. Use when the user asks "how many API calls do I have?", "what plan am I on?", "what graphs can I access?", or similar account questions. Returns live data — not cached from session start. (User account status & credit balance lookup.)

Catalog and discover all expert personas available on Fodda across 4 categories: living Human Agents (practitioners with Human Agent models), C-Suite Agents (corporate executive lenses grounded in earnings), Classic Agents (historical thinkers), and Synthetic Analysts (domain specialists). Use when exploring available experts, finding who to consult for a specific topic, or inspecting available specialist deliverables. Sibling routing: This tool lists and filters agents; to initiate an interactive conversation with a chosen specialist, call consult_human_agent (for living experts) or consult_analyst (for synthetic/classic/C-suite personas); to search experts by query match, use find_expert. Returns an array of agent profiles with analyst_id, canonical_slug, name, role, agent_class, topics, consult_tool, and commissionable offerings.

Step 1 (session initializer) for Fodda Human Agent onboarding: verifies credentials, checks for existing active or in-progress onboarding sessions, and initiates profile creation for a living expert. Use when an expert asks to create their digital twin or join the Fodda network (e.g., 'I want to build my Human Agent', 'How do I onboard as an expert on Fodda?'). Supports resuming interrupted onboarding sessions automatically. Sibling routing: Use begin_expert_onboarding to kick off or resume onboarding; use submit_basic_info to save initial profile credentials and accept terms; for BYO-MCP onboarding, pass byoMcp: true. Returns onboarding state, current step, and guided next actions.

Explore and brainstorm around a topic using knowledge graph connections. Unlike search (which finds what matches), this tool discovers what CONNECTS — adjacent trends, unexpected cross-domain links, key brands, and geographic hotspots. Use when the user wants to brainstorm, explore adjacencies, find inspiration, or understand the landscape around a topic. Returns a structured brainstorm map with territories to explore.

Use when auditing a single brand's 360-degree intelligence footprint: evaluates brand health, trend footprint, competitor co-occurrences, Google Trends velocity, and Wikipedia pageviews across 100+ knowledge graphs. Mandatory entry point whenever a prompt names a specific commercial brand or company (e.g., 'Audit Nike's market footprint', 'How is Lululemon performing across consumer trends?'). Sibling routing: Use search_graph for topic- or category-level discovery (e.g. 'athleisure trends'); use get_company_earnings for detailed SEC financial statements and earnings Q&A; use brand_tracker when focusing on a specific brand's multi-graph presence, peer competitor matrix, and cultural momentum. Returns a comprehensive brand dossier containing trend_footprint, co_occurring_brands (competitor matrix), search_interest metrics, evidence_velocity, and optional pre-rendered interactive widget_html.

Returns Fodda's capabilities, semantic research workflows, Human Agents (Human Agents), and pricing. Call this for any question about what Fodda can do, how experts work, available research recipes, or pricing. Sibling routing: Use get_capabilities to discover platform features and workflow recipes; use list_graphs to list specific data graphs; use list_analysts to discover expert personas. Returns platform overview, active specialist counts, workflow descriptions, and published offerings.

The canonical per-ticker earnings source for 517 covered companies. - view='snapshot' (default): Full quarterly record including headline, financial metrics, analyst sentiment, strategic activity, CEO quotes, and top Q&A entries. - view='qa': Itemized analyst Q&A session with verbatim analyst questions, analyst names, analyst firms, executive responses, and response directness. USE THIS VIEW whenever the user asks who asked a question, for analyst names/firms, or for verbatim call quotes. - view='history': Narrative quarterly timeline. - view='compare': Side-by-side comparison of 2-5 tickers.

Step 4 in standard Human Agent onboarding: locks the expert's selected 2-3 core expertise themes and triggers automated generation of their tailored voice interview questionnaire. Prerequisite: get_detected_themes must be reviewed first to select themes and flag any inaccurate public findings. Workflow continuation: Successfully confirming themes advances the onboarding session to schedule_interview for the 5-10 minute voice interview. Sibling routing: Use get_detected_themes to inspect detected themes before confirming; use finalize_byo_mcp_onboarding instead if following the Bring-Your-Own-MCP pathway. Returns confirmation status, generated interview question list, and next-step instructions for scheduling the interview.

Use when commissioning an autonomous multi-pass research report synthesizing cross-graph trends, corporate disclosures, and expert perspectives into a comprehensive executive narrative (e.g., 'Generate a comprehensive research briefing on the circular fashion economy', 'Deep dive into retail media network adoption'). Runs asynchronously; execution returns a job_id that must be polled to completion using check_research_status. Sibling routing: Use search_graph for fast, single-turn trend lookups; use get_intelligence_dossier for immediate synchronous multi-source dossier assembly; use deep_research_topic when you need an autonomous, multi-step deep dive. Returns job confirmation, estimated processing time, and job_id for status tracking.

Poll a deliverable commissioned with request_deliverable. Pass the job_id from that response. Returns the current status ("working" | "completed" | "failed") and, once completed, the artifact links to present to the user. Polling is free. Deliverables typically take a few minutes — poll every ~15–30s. (Expert deliverable job status poll.)

Step 4a in standard Human Agent onboarding: retrieves detected expertise themes and verified public findings for expert review before locking. Use after submit_expertise_analysis to present themes and cited public articles to the expert for approval. Prerequisite: submit_expertise_analysis must be completed first. Workflow sequence: Present the findings to the expert to flag inaccurate items, select 2-3 core themes, and then proceed to confirm_themes. Returns an array of detected theme objects and verified research findings with citation URLs.

Vector-similarity cross-domain trend discovery: uncovers non-obvious, structurally parallel trend patterns across disparate industries without relying on direct editorial links. Use for lateral ideation, analog benchmarking, or cross-vertical innovation (e.g., 'Find cross-industry parallels to automotive subscription models in hospitality', 'What trends mirror luxury resale mechanics in healthcare?'). Sibling routing: Use search_graph for direct keyword/topic trend matching; use get_neighbors to traverse direct explicit graph edges (e.g. brands, case studies); use brainstorm_topic for conceptual ideation; use discover_adjacent_trends specifically for structural analogies and cross-domain pattern matching. Returns ranked parallel trend nodes, vector similarity scores, and cross-vertical rationale.

Search PSFK-curated domain graphs (travel & hospitality, retail, tech, beauty, fashion, sports, food & beverage) for trend intelligence with bundled evidence. No graph ID needed — searches all 7 live domain graphs in parallel. Returns expert-curated trends with bundled evidence including brand case studies, statistics, executive quotes and analysis with source attribution. Sibling routing: Use get_domain_intelligence when exploring broad consumer culture across the primary retail/consumer domains without needing a graphId; use search_graph to query the broader catalog of 100+ specialist partner graphs; use get_report_intelligence for published enterprise whitepapers; use brand_tracker if querying a specific company or brand name.

Evidence citation lookup: retrieves primary evidence articles, publisher case studies, data points, and source URLs that validate a specific trend node. Prerequisite: Requires a valid for_node_id and graphId obtained from a prior search_graph or get_neighbors call. Use when backing a specific trend with original reporting, verifying publisher citations, or extracting concrete brand examples. Sibling routing: Use search_graph first to discover trends and retrieve their node_id; use get_evidence to inspect the underlying source citations for that node; use get_node for high-level metadata; use search_statistics for standalone numbers. Returns an array of evidence items with title, publication, publication_date, snippet, and source_url.

Step 3 (final step) in the Bring-Your-Own-MCP (BYO-MCP) onboarding pathway: persists the completed expert profile and registers the live external MCP grounding in the Fodda Human Agent registry. Profile is saved once terms are accepted; MCP details are verified and submitted for review here. Prerequisites: submit_basic_info and submit_mcp_source must be completed prior to finalizing; terms must be accepted. Workflow distinction: This tool is strictly for BYO-MCP experts whose knowledge is grounded in an external MCP server—do NOT use for standard graph-backed experts (who proceed through confirm_themes and schedule_interview). Side effects: Creates an active pending-review Human Agent record in Airtable and provisions consultation routing. Returns submission status, analystId, and confirmation details.

Use when searching for market trends, category dynamics, and consumer shifts across 100+ curated knowledge graphs. Returns trends with cited evidence, lifecycle stage (emerging/building/mature), and structured next_moves. Sibling routing: Use search_graph as the default starting tool for broad trend discovery; use search_statistics if you specifically need numerical data points and growth rates; use search_insights if you only need qualitative quotes and expert commentary; use get_domain_intelligence to search all 7 core PSFK industry verticals simultaneously; use brand_tracker if querying a specific company or brand name.

Catalog and discovery layer for Fodda knowledge graphs: lists all accessible knowledge graph IDs, topic vertical tags, author credentials, node counts, and signal metrics. Use first when discovering available domain libraries or when you need a valid graphId parameter for graph-scoped queries (e.g. "What knowledge graphs do you have on retail or sustainability?"). Sibling routing: Use list_graphs to inspect graph metadata and available sources; use search_graph to query trend content across graphs; use list_analysts to discover specialist Human Agents. Returns an array of graph profiles with graphId, name, topics, signal_count, and author.

Narrative and qualitative evidence layer only: returns expert quotes, editorial analysis, and strategic perspectives from named strategists and industry leaders across all graphs, with source attribution and parent trend context. Sibling routing: Use search_insights when you need authoritative quotes and qualitative commentary; use search_statistics for numerical data points and market sizing; use consult_human_agent to speak interactively with an expert. Does NOT return raw statistics or market sizing.

Certified multi-source evidence dossier generator: synthesizes domain knowledge graph case studies, consulting report findings (PwC, McKinsey, DHL), expert quotes, and live macroeconomic data (FRED, BLS, Census) into a comprehensive intelligence dossier in a single synchronous call. Use when preparing client pitch decks, executive briefings, or comprehensive topic overviews (e.g., 'Build an evidence dossier on sustainable luxury packaging'). Optionally generates an executive 4-beat journalistic briefing (include_editorial: true). Sibling routing: Use get_intelligence_dossier for instant, single-turn multi-source synthesis; use deep_research_topic for multi-pass autonomous background research; use search_graph for individual trend queries. Returns synthesized strategic narrative, categorized evidence items, quantitative metrics, and contrarian adversarial points.

List unique taxonomy entity values (brands, locations, technologies, audiences) or retrieve the complete table of contents for small, discrete report graphs (<40 trends, e.g. DHL, PwC, or Mintel radars) where search may return partial results. Returns up to 100 unique values ordered alphabetically. Sibling routing: DO NOT use this tool for thematic or topic research (e.g. 'find AI trends' or 'what trends are in tech?') on large domain graphs (tech, retail, cpg), because it caps at 100 items alphabetically without evidence or signals; use search_graph or get_domain_intelligence instead for semantic trend retrieval. (Graph taxonomy label value lookup.)

Use when a user asks to track a topic, monitor a competitor, or set up, list, update, pause, resume, or cancel recurring weekly/daily intelligence briefings delivered via email (e.g., 'Track sustainability shifts in luxury fashion every Monday', 'Pause my weekly retail digest'). Actions: 'create' registers a new schedule (requires query, email, cadence); 'list' returns all active schedules for the user; 'update' modifies query/cadence for schedule_id; 'pause' and 'resume' temporarily halt or restart delivery; 'cancel' permanently removes the schedule. Returns confirmation status, schedule_id, active frequency, and next delivery date.

Discover what's connected to a specific trend — related brands, technologies, locations, and cross-domain links. Returns curated editorial connections between trends. Use after search_graph to map the territory around a trend, find which brands are connected, or understand cross-domain relationships. Requires node_id from a prior search_graph result. (1-hop graph relationship traversal from seed node in Neo4j.)

Get the full profile of a specific trend — detailed description, lifecycle stage (emerging/building/mature), signal strength, geographic scope, and all properties. Use when you need deeper detail on a single trend after search_graph returned a summary. Requires node_id from a prior search_graph result. (Direct Neo4j node metadata lookup by ID.)

Status and progress checker for Human Agent onboarding sessions: retrieves current milestone, active stage, and pending tasks. Supports holding connection (waitForResearch: true) for up to 25 seconds while background research completes. Use when monitoring onboarding progress or checking whether asynchronous background research has finished. Sibling routing: Use get_onboarding_status for expert onboarding tracking; use check_research_status for topic deep research; use check_deliverable_status for commissioned documents. Returns session status, completed steps, research_state, next_tool, and plain-English next_action instructions.

Published corporate research and market forecast layer: searches industry report knowledge graphs (DHL, PwC, Unilever, Jack Morton, and specialist research firms) for published forecasts, projections, and whitepaper findings. Returns an executive 5-pillar analyst briefing by default with cross-graph validation. Does NOT return living consumer domain trends (use get_domain_intelligence) or standalone data points (use search_statistics). When the query names a specific company or brand, brand_tracker is the entry point. No graph ID needed.

Polling and retrieval layer for deep research jobs: checks execution status and retrieves the completed research briefing report. Call after initiating deep_research_topic—poll every 10 seconds until status transitions to COMPLETE or FAILED. Sibling routing: Use check_research_status for deep_research_topic jobs; use check_deliverable_status for analyst documents; use check_supplemental_status for supplemental data queries. Returns execution status (IN_PROGRESS, COMPLETE, FAILED), completion percentage, and the full Markdown briefing report upon completion.

Step 5 (final stage) in standard Human Agent onboarding: schedules or immediately dispatches a 5–10 minute voice interview with the Fodda AI interviewer on Google Meet to capture conversational tone, voice cadence, and contrarian perspectives. Prerequisite: confirm_themes must be completed first to generate the interview questionnaire. Sibling routing: This tool finalizes standard graph-backed expert onboarding; for Bring-Your-Own-MCP onboarding, use finalize_byo_mcp_onboarding instead (which does not require a voice interview). Returns meeting link, calendar confirmation, and next steps for going live.

Forward user feedback, feature requests, complaints, or exit reasons to the Fodda team via email and Slack. Call this whenever a user shares feedback — including when they want to leave, report a problem, or suggest an improvement. (User feedback logging.)

Create a free Fodda Base account (100 API calls/month across ALL knowledge graphs) and send a confirmation email. GUARDRAIL: only call this AFTER the user has explicitly provided their email and asked to create an account — never sign someone up proactively or with an email inferred from earlier context. Can also pass profile fields (name, job_title, company). (User registration and free tier provision.)

[Deprecated: Prefer search_graph for topic-routed multi-graph searches, or consult_human_agent / consult_analyst for direct strategist and thinker consultation.] Proprietary strategist frameworks and niche domain intelligence layer: searches specialist knowledge graphs curated by domain strategists, newsletters, and boutique studios (culture, youth trends, commerce, media). Returns specialist frameworks and analysis; does not return broad consumer domain trends (use get_domain_intelligence) or standalone statistics (use search_statistics). When the query names a specific company or brand, brand_tracker is the entry point. No graph ID needed.

Quantitative statistics and hard numbers retrieval layer: queries curated knowledge graphs for empirical metrics, survey percentages, market sizes, adoption rates, and compound annual growth rates (CAGR) linked to parent trends. Use when looking specifically for data points, percentages, and numerical proof to substantiate arguments (e.g., 'What percentage of Gen Z consumers prefer clean beauty?', 'Find market size data for sports tourism'). Sibling routing: Use search_statistics for numerical data points; use search_insights for quotes and narrative editorial commentary; use get_domain_intelligence for full trends with bundled case studies; use get_supplemental_context for live government/macro statistics. Does NOT return general narrative prose.

User search scope preference manager: toggles inclusion or exclusion preferences for specific knowledge graphs (e.g., 'Don't include the sports graph in my searches', 'Re-enable the beauty graph'). Preferences persist across sessions in the user profile. Sibling routing: Use toggle_graph_preference to set permanent search filtering; use update_user_profile for general account context; pass explicit graphs or graphId directly on search_graph for one-off query overrides. Returns updated user graph preference state and confirmation status.

Extract clean text content from any URL. Use this when a user shares a link (competitor site, news article, client brief, trend report) and wants to cross-reference it against Fodda knowledge graphs. Returns structured text ready for analysis.

Save the user's research profile to improve the relevance of future responses. Call this after you understand the user's role, industry, and research needs. The profile persists across sessions — you only need to set it once, then update if their focus changes. Write BEHAVIORAL INSTRUCTIONS, not a bio. Format: one sentence of identity (who they are and how they use Fodda), then numbered directives that change how you synthesize and frame responses. Include: what evidence to prioritize, how to frame conclusions, geographic needs, and output structure preferences. Max 2000 chars per field. (User persona context update.)

Discovers consumer trends that have been empirically corroborated by corporate earnings reports in Fodda's cross-validation pipeline. Connects high-level cultural trend signals with corroborating executive statements and analyst Q&A from quarterly 10-Q/10-K earnings calls. Use when you need proof that an emerging consumer trend is impacting corporate P&L and executive strategy (e.g., 'Show consumer trends validated in recent beauty earnings calls'). Sibling routing: Use search_graph for pure cultural/consumer trend discovery; use get_company_earnings for full single-ticker filings; use get_validated_trends when you specifically need the bridge between consumer trends and corporate earnings validation. Returns validated trend records with corroborating company tickers, validation types, analyst themes, and executive quote excerpts.