Skip to main content
Glama

Hyperresearch Answers

Cited answer from filings, economic data, patents and the web

data_and_web_answer

Answers a factual question from the record itself — SEC EDGAR company filings, FDIC bank records, FEC campaign-finance and lobbying disclosures, USPTO patents and trademarks, FRED and Census economic series, ClinicalTrials.gov registrations and openFDA records, NOAA weather and climate records, scholarly databases and books, market data and posts on X, among 120+ structured public data sources, plus the live web. Returns a short answer checked word for word against quoted sources with their URLs, or an honest not-found. Suited to market size and growth, competitors, company facts and finances, the economy, pricing, statistics, regulations and recent events. Ask in plain language, one question per call; split a broad brief into several asks. Mode fast reads the open web only; for filings, government statistics and registries use standard or auto. Hard questions can take 1-5 minutes. If the result says pending, call get_pending_answer with the id; you will not be charged twice. Calls start at 2¢; each call is charged to the user's prepaid Hyperresearch Answers balance.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "question": "<question>"
    },
    "name": "data_and_web_answer"
  }
}

Input Schema

NameRequiredDescriptionDefault
modeNoHow hard to look: fast (2¢, no escalation, typically about 2 s), standard (10¢, 20¢ if it has to dig deeper; the default, up to about 3 min) or max (50¢, numbers confirmed on two sites, up to about 5 min). fast answers from the open web only and does not query structured data sources; for a figure from a company filing, a government statistic or a registry (SEC EDGAR, FRED, Census, FDA, FEC, USPTO, ClinicalTrials.gov), use standard or auto. auto: we pick fast, standard or max for the question and charge the price of the mode that ran; the result says which mode ran. A question that needs a written report also gets a suggestion to run Light or Deep research (only suggested, never started). An honest not-found costs 20% to 100% of the call's price, rounded up to the cent (1¢ to 2¢ on fast; up to 20¢ for a standard ask that digs deeper). A paywalled article the call reads may add a surcharge of 1.5 times the article's licence cost (licence price plus transaction fee), rounded up to the cent. `deep` is accepted as the old name for max.standard
questionYesOne factual question, asked the way you would ask a researcher.
exclude_textNoShort phrases that must not appear in the returned text.
include_textNoShort phrases to look for in the returned text. Standard search keeps a result with any of them; max search and ask need all of them.
max_cost_centsNoThe most this call may cost, in cents. Below the mode price the call is refused. For standard, below 20 it never digs deeper and costs at most 10¢.
exclude_domainsNoNever sources on these domains (or subdomains).
include_domainsNoOnly sources on these domains (and subdomains), e.g. ["sec.gov"].
end_published_dateNoOnly sources published on or before this ISO date (YYYY-MM-DD). Undated sources are left out when a date is set.
start_published_dateNoOnly sources published on or after this ISO date (YYYY-MM-DD). Undated sources are left out when a date is set.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Added

TDQS

A4.5/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations only tell the agent this is non-read-only, open-world, non-idempotent; the description adds the details that actually matter: 1-5 minute latency, per-call pricing from 2¢, billing to a prepaid balance, the not-found partial refund (20-100% of price), a paywall surcharge, and the no-double-charge guarantee on pending retries. That is substantial behavior beyond structured fields.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The opening sentence front-loads what the tool does and the citation guarantee, and subsequent sentences carry distinct payloads (modes, latency, pending handling, cost). It is dense and long, with some duplication of the mode/cost rules already in the schema, which keeps it off a 5.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 9-parameter, no-output-schema tool with metered billing, the definition covers return shape, latency, cost, refund semantics, and the pending workflow. The remaining gap is minor: no guidance on failure/error behavior beyond not-found, and no statement on whether repeated identical questions are re-charged.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100% and the mode enum, costs, and include/exclude/date constraints are all documented in the schema itself, so the baseline is 3. The description mostly restates the fast-vs-standard rule and adds only a light hint on the question parameter ('ask in plain language, one question per call'); the six filtering parameters are never mentioned in the prose.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a precise verb and resource ('Answers a factual question from the record itself'), enumerates the source families, and pins down the output shape ('a short answer checked word for word against quoted sources with their URLs, or an honest not-found'). The pending-result handoff to get_pending_answer makes the boundary with that sibling explicit, and the 'short cited answer' framing separates it from a raw search tool.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives concrete routing rules: 'split a broad brief into several asks', 'one question per call', 'Mode fast reads the open web only; for filings, government statistics and registries use standard or auto'. It also names the follow-up tool and condition ('If the result says pending, call get_pending_answer with the id'). Both when-to-use and when-not-to-use are covered.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.