mcp.agentii.ai
https://mcp.agentii.ai/mcp
mcp_partial Not production-safe Score 89/100 http no auth
At a glance
| Endpoint | https://mcp.agentii.ai/mcp |
|---|---|
| Source | https://raw.githubusercontent.com/agentii-ai/agentii-investment-intelligence/main/plugins/vertical-plugins/equity-research-core/.mcp.json |
| Transport | http |
| Tools | 0 |
| Auth | none |
| First seen | 2026-08-29 15:33:38 |
| Last checked | 2026-11-16 16:54:06 |
Observed reliability
| 7-day uptime | 99% |
|---|---|
| 24-hour uptime | — |
| Handshake success | 100% |
| Latency p50 / p95 | 811 / 1267 ms |
| Last successful check | 2026-11-16 16:29:21 |
Measured by Arclan's validator via repeated live MCP handshakes. How this is computed →
Tools (0)
search_documents— **Layer 1 — Document Discovery.** Use FIRST to find which filings exist for a ticker before any page-level operation. Returns document metadsearch_sec_filings— **Layer 1 — Document Discovery.** Use FIRST to discover SEC filings by form_type, date, or ticker. For US companies: use form_type=["10-K","search_xbrl_facts— Search XBRL financial facts. Default returns only is_primary=true facts (most authoritative source per concept+period: 10-K/20-F > 10-Q > 8-list_xbrl_concepts— List distinct XBRL concepts available for querying. Use BEFORE search_xbrl_facts to avoid concept hallucination.get_company_financials— Get company financial overview: XBRL filing metadata, highlights, metrics in one call. Aggregates across multiple XBRL tables. Pass fiscal_ysearch_companies— Search companies by ticker, name, exchange or medical classification (spec 071): business_segment, business_model, oncology_focus, therapeutsearch_universe_drugs— Search the drugs universe (spec 049). Default returns ingredient-group level (gold.drug_ingredients — the searchable units with derived profsearch_universe_devices— Search the devices universe (spec 049). Default returns product-code-grouped summaries; pass flat=true for raw submission rows (gold.devicesget_company_drugs— Get the drugs a company develops (spec 049) — ingredient groups with applications, via the universe resolution. Key by ticker, company_id, oget_company_devices— Get the devices a company manufactures (spec 049) — grouped by product code via the universe resolution. Key by ticker, company_id, or compaget_upcoming_pdufa— List upcoming FDA/PDUFA action dates from the company-owned SEC calendar (spec 049) — all future events by default, date-ascending; period-oget_pdufa_decision— PDUFA decision history for a drug/ticker/application (spec 049) — target date, decision date, outcome, verification, and the superseded dateget_device_decision— Forward-looking device decisions (PMA/De Novo/510(k)/HDE) the company disclosed in SEC filings (spec 049) — goal dates, verification, outcomget_company_profile— Get enriched company profile with fiscal calendar, SEC filing counts, XBRL filing count, and coverage dataget_company_classification— Get a company's medical classification (5 company axes: business_segment, business_model, oncology_focus, therapeutic_focus, development_staget_company_hierarchy— Get a company's ownership graph (spec 071): parent chain, subsidiaries with source and evidence, and the count of FDA drug applications / desearch_company_assets— Search company assets by label (spec 071) — asset kind, therapeutic area, oncology classification, therapeutic modality, development stage,peer_of— Bounded, explainable peer set for a medical company (spec 071): peers share at least 2 of 9 dimensions (5 company axes plus asset_kind, thersearch_earnings_calendar— Search earnings calendar events with EPS actual/estimate/surprise. Each row carries `source_snapshot_date` (the FMP snapshot the row came frlist_upcoming_earnings— List upcoming earnings report dates within next N days (default 30, max 90). Grouped by report_date.list_coverage— List per-source ticker coverage, record counts, latest data dates, and freshness tiers.get_ticker_coverage— Get coverage and freshness metadata for a single ticker across all source types.list_domains— List available knowledge domains from domain_registry (9 rows).get_company_fiscal_calendar— Get the issuer's declared fiscal chain from XBRL dei facts — actual periods, period_id, record_status and gaps. No December default and no sresolve_period— **Temporal resolver.** Turn a fiscal/calendar phrase (FY2026 Q3, Q4 FY2026, CY2025 Q3, this/last/next quarter) into the issuer's declared pelocate_in_source— **Deterministic locator.** Fuse the pages you chose from the map (`map_picks`) with a keyword channel over the document, server-side (RRF k=locate_figures— **Figure locator.** Find the pages of a document that carry the printed figures you name — as printed (`1,234`) or as values to render (`123search_transcript_turns— **Transcript turns.** Search one earnings-call transcript's speaker turns — who said it, in which section (prepared or Q&A), and the page spfind_evidence— **Evidence plan executor.** Compose the page locator and figure locator for one evidence need (or a free-text question) and return candidatesearch_unified— Cross-source search by ticker + keyword. Results grouped by source_type.list_sources— **Layer 1 — Document Discovery.** Discover available data sources for agentic retrieval by ticker, year, source_type. Sources include citatiread_source_outline— **Layer 2 — Page Map (Lightweight).** Use AFTER document discovery (search_documents/search_sec_filings) and BEFORE read_source_pages to scaread_source_deep_outline— **Layer 2 — Page Map (Deep).** Use AFTER read_source_outline when table_titles, drivers, metrics, or views are needed to disambiguate betweesearch_keyword_in_source— **Layer 2 — Keyword Filter.** Use AFTER read_source_outline and BEFORE read_source_pages. Preferred: pass ticker + citation_id from search rread_source_pages— **Layer 3 — Deep Read.** Use AFTER read_source_outline to select relevant pages. Preferred: pass ticker + citation_id + pages (e.g., LLY + sverify_quote— **The check to run immediately BEFORE you write a quotation.** Pass the sentence you are about to write, with the page you are about to citebatch_search— Execute up to 20 parallel sub-queries. Each sub-query specifies endpoint + params. Results keyed by index. Partial failure supported.search_cross_period— **Multi-Period Search.** Resolves each period through resolve_period first, then searches the issuer's documents. Pass `period_ids` from resget_statement— Get complete multi-period financial statement for an SEC filing. Returns dual-format: structured rows (concept, label, values by period) + rget_calculation_tree— Get accounting formula tree for an SEC filing from XBRL calculation linkbase. Returns nested arcs (parent_concept, child_concept, weight) grvalidate_calculation— Cross-validate XBRL calculation tree against reported fact values for an SEC filing. Computes SUM(child * weight) for each parent concept anget_statement_structure— Get report line-item tree from XBRL presentation linkbase for an SEC filing. Returns nested parent/child concept relationships with labels,get_segment_data— Get segment-level financial data grouped by XBRL dimensional axes (business unit, product, geography). Returns revenue, operating income, asget_peer_comparison— Get peer company financial metrics for cross-company comparison. Finds sector peers via gold.companies and fetches metrics via gold.xbrl_facsearch_institutional_holdings— Search institutional 13F holdings for a ticker — which institutions hold it (top-10 holders per ticker + full portfolios of 97 mega-institutsearch_insider_trades— Search Form-4 insider trading activity for a ticker — purchases/sales/awards by officers, directors, and 10% owners, with transaction price,get_eps_growth— Get 5-year EPS history with 3yr/5yr CAGR and consensus LTG estimate. Computes growth rates from historical EarningsPerShareDiluted facts andget_financial_ratios— Get pre-computed financial ratios for a ticker. Returns profitability (ROE, ROA, ROIC, margins), liquidity (current/quick/cash ratios), levesearch_investment_cases— Search historical investment cases from the gold.investment_cases table. Cases are structured narratives of specific completed investment epget_investment_case— Get a single investment case by case_id with full detail (thesis, timeline, evidence, bear case, outcome, lessons learned). Use AFTER searchsearch_investment_strategies— Search reusable investment strategies from the gold.investment_strategies table. Strategies are institutional-grade methodologies classifiedget_investment_strategy— Get a single investment strategy by strategy_id or citation_id. Returns full detail including body (selection_criteria, position_sizing, rissearch_technical_setups— Search technical entry/exit setups from the gold.technical_setups table. Setups are specific execution rules with concrete price/indicator vget_technical_setup— Get a single technical setup by setup_id or citation_id. Returns full detail including entry_setup, confirmation_signals, exit_rules, invalisearch_by_analogue— Cross-cutting analogue search across BOTH gold.investment_cases AND gold.investment_strategies. Given a market_regime, event_type, or companlist_related_cases— List all investment cases related to a given case_id. Returns cases linked via the related_cases column in gold.investment_cases. Use to tralist_related_entries— List all knowledge entries related to a given entry_id. Returns entries linked via the related_entries column in gold.knowledge_entries. Usesearch_knowledge_entries— Search analytical frameworks and methodologies from the gold.knowledge_entries table (K1-K8: Dalio Debt Cycle, Minsky Hypothesis, Mauboussinget_knowledge_entry— Get a single analytical framework by entry_id with full detail (body, related_entries, provenance). Use AFTER search_knowledge_entries to drretrieve_and_contextualize— Composite tool: parallel-fetches knowledge entries, investment cases, strategies, analogues, and technical setups for a ticker+domain combin
Validation history
| Checked | Reachable | Handshake | Latency | Score | State |
|---|---|---|---|---|---|
| 2026-11-16 16:54:06 | ✓ | ✓ | 230 ms | 55 | mcp_partial |
| 2026-11-16 16:29:21 | ✓ | ✓ | 797 ms | 90 | active |
| 2026-11-16 16:08:28 | ✓ | ✓ | 858 ms | 90 | active |
| 2026-11-16 15:40:52 | ✓ | ✓ | 788 ms | 90 | active |
| 2026-11-16 15:24:12 | ✓ | ✓ | 801 ms | 90 | active |
| 2026-11-16 15:09:09 | ✓ | ✓ | 852 ms | 90 | active |
| 2026-11-16 14:54:08 | ✓ | ✓ | 964 ms | 90 | active |
| 2026-11-16 14:36:26 | ✓ | ✓ | 603 ms | 90 | active |
| 2026-11-16 14:18:54 | ✓ | ✓ | 965 ms | 90 | active |
| 2026-11-16 14:03:42 | ✓ | ✓ | 640 ms | 90 | active |
| 2026-11-16 13:48:31 | ✓ | ✓ | 793 ms | 90 | active |
| 2026-11-16 13:33:20 | ✓ | ✓ | 709 ms | 90 | active |
| 2026-11-16 13:16:12 | ✓ | ✓ | 811 ms | 90 | active |
| 2026-11-16 13:01:00 | ✓ | ✓ | 630 ms | 90 | active |
| 2026-11-16 12:46:00 | ✓ | ✓ | 1018 ms | 86 | active |
Query this programmatically
Agents can pull this record live over the MCP protocol or REST:
get_server("https://mcp.agentii.ai/mcp") # via the Arclan MCP endpoint https://arclan.ai/mcp
GET https://arclan.ai/api/registry/mcp-agentii-ai-mcp # JSON