What MCP changes, what it does not change, and how financial data differs from market data.MCP 改变什么、不改变什么，以及金融数据与市场数据的区别。
Production guide for financial data MCP
金融数据 MCP 生产指南
Financial Data MCP Server for AI Agents 面向 AI Agent 的金融数据 MCP 服务器
Learn how to connect agents to filings, company fundamentals, earnings, macroeconomic series, and financial news through typed MCP tools—with validation, provenance, access controls, and an auditable QVeris workflow. 了解如何通过类型明确的 MCP 工具，让智能体访问监管文件、公司基本面、财报、宏观经济序列和金融新闻，并保留验证、来源、访问控制与可审计的 QVeris 工作流。
6
canonical QVeris MCP tools个 QVeris 标准 MCP 工具
2
local and hosted transports种本地与托管传输方式
7
financial data control checks项金融数据控制检查
1
discover-to-audit workflow套发现到审计工作流
Governed financial data受控金融数据
Filings, fundamentals, earnings, macro series, and research context.
监管文件、基本面、财报、宏观序列与研究上下文。
Typed MCP access类型明确的 MCP 访问
Discover schemas before an agent spends credits or calls a provider.
在智能体消耗积分或调用供应商前发现并检查参数结构。
Evidence after every call每次调用都有证据
Keep source, timestamp, period, units, execution ID, and settlement records.
保留来源、时间戳、报告期、单位、执行 ID 与结算记录。
TL;DR
A financial data MCP server is a governed tool boundary—not a data source 金融数据 MCP 服务器是受控工具边界，而不是数据源本身
A financial data MCP server exposes financial capabilities to an MCP-compatible agent through schema-defined tools. The server can wrap company fundamentals, SEC filings, earnings events, transcripts, macroeconomic series, and news APIs while keeping provider credentials and execution logic outside the model context. 金融数据 MCP 服务器通过结构化工具向兼容 MCP 的智能体暴露金融能力。它可以封装公司基本面、SEC 文件、财报事件、电话会记录、宏观经济序列和新闻 API，同时把供应商凭据与执行逻辑留在模型上下文之外。
MCP standardizes how the agent discovers and invokes a tool; it does not make the underlying data accurate, licensed, timely, or comparable. Production teams still need source provenance, period and unit checks, entitlement enforcement, rate limits, and golden-query validation. MCP 规范了智能体如何发现和调用工具，但不会自动保证底层数据准确、已授权、及时或可比。生产团队仍需处理来源追踪、报告期与单位校验、权限执行、速率限制和基准查询验证。
Use this guide if you need to本指南适合以下需求
1
Connect Claude, Cursor, or another MCP client to provider-backed financial tools.把 Claude、Cursor 或其他 MCP 客户端连接到供应商支持的金融工具。
2
Build a filing, fundamental, earnings, macro, or research agent with auditable calls.构建具有可审计调用的文件、基本面、财报、宏观或研究智能体。
3
Choose between a direct API, a custom MCP server, and a capability routing layer.在直接 API、自建 MCP 服务器和能力路由层之间做选择。
4
Validate data before it reaches an investment memo, alert, or downstream model.在数据进入投资备忘录、告警或下游模型前完成验证。
Guide Map指南目录
What this financial data MCP guide covers本金融数据 MCP 指南涵盖什么
Host, client, server, provider, audit layers, and the metadata every result needs.Host、客户端、服务器、供应商、审计层，以及每个结果必须携带的元数据。
Local stdio and hosted HTTP options, verification, discovery, inspection, probe, and call.本地 stdio 与托管 HTTP 方案，以及验证、发现、检查、探测和调用。
A practical earnings-research sequence that combines filings, facts, events, and macro context.结合文件、财务事实、事件与宏观背景的实际财报研究流程。
Golden queries, freshness, units, revisions, permissions, licensing, and audit evidence.基准查询、时效、单位、修订、权限、许可与审计证据。
When a direct provider, custom server, or QVeris routing layer is the better fit.何时直接供应商、自建服务器或 QVeris 路由层更合适。
Definition定义
What is a financial data MCP server?什么是金融数据 MCP 服务器？
It is an MCP server that exposes narrowly scoped financial-data operations as discoverable, schema-defined interfaces an AI application can invoke. 它是一种 MCP 服务器，把范围明确的金融数据操作暴露为可发现、具有参数结构、可由 AI 应用调用的接口。
Examples include resolving a ticker to an entity, retrieving a filing, reading company facts, fetching an economic series, or searching financial news. Inputs and outputs should be typed and bounded.例如把股票代码解析为公司实体、获取监管文件、读取公司事实、获取经济序列或搜索金融新闻。输入和输出都应类型明确且范围受限。
Resources can expose read-only schemas, metric definitions, filing indexes, provider documentation, or approved research templates without treating every read as an action.Resources 可提供只读的数据结构、指标定义、文件索引、供应商文档或已批准的研究模板，而不把每次读取都视为一次操作。
A prompt can standardize an earnings brief or company comparison, but the evidence must still come from tools and resources with explicit source metadata.Prompt 可以标准化财报摘要或公司比较，但证据仍应来自带有明确来源元数据的 tools 与 resources。
Important boundary: MCP is the interaction protocol, not the financial-data license, warehouse, semantic layer, or truth system. The authoritative source may still be an SEC endpoint, a macroeconomic database, an exchange-licensed feed, or a commercial fundamentals provider. 重要边界：MCP 是交互协议，不是金融数据许可证、数据仓库、语义层或事实系统。权威来源仍可能是 SEC 接口、宏观经济数据库、交易所授权行情或商业基本面供应商。
Data Scope数据范围
Financial data and market data are related, but not identical金融数据与市场数据相关，但并不相同
This page focuses on research and financial context. Low-latency quotes, trades, bars, order books, exchange sessions, and streaming belong in the separate market data MCP server guide. 本页聚焦研究与金融上下文。低延迟报价、成交、K 线、订单簿、交易时段和流式数据应参考单独的市场数据 MCP 服务器指南。
| Category类别 | Typical operations典型操作 | Critical metadata关键元数据 | Primary risk主要风险 |
|---|---|---|---|
| Entity and security master实体与证券主数据 | Resolve ticker, CIK, ISIN, exchange, share class解析股票代码、CIK、ISIN、交易所和股权类别 | Entity ID, effective dates, exchange实体 ID、生效日期、交易所 | Wrong-company joins错误公司关联 |
| Filings and disclosures监管文件与披露 | Find 10-K, 10-Q, 8-K, retrieve document or XBRL facts查找 10-K、10-Q、8-K，获取文档或 XBRL 事实 | CIK, form, accession, filed date, periodCIK、表格类型、accession、提交日期、报告期 | Mixing filing and fiscal dates混淆提交日期与财务报告期 |
| Fundamentals公司基本面 | Revenue, margins, cash flow, balance sheet, ratios收入、利润率、现金流、资产负债表与比率 | Taxonomy, units, currency, period, restatement分类标准、单位、币种、报告期、重述状态 | False cross-company comparability错误的跨公司可比性 |
| Earnings and transcripts财报与电话会记录 | Calendar, estimates, reported values, transcript retrieval财报日历、预期值、实际值与电话会记录获取 | Event time, fiscal quarter, speaker, source rights事件时间、财季、发言人和来源权利 | Using preliminary data as final把初步数据当作最终数据 |
| Macroeconomic series宏观经济序列 | Search series, observations, releases, vintage dates搜索序列、观测值、发布与历史版本日期 | Frequency, units, seasonal adjustment, vintage频率、单位、季调状态、数据版本 | Look-ahead bias after revisions修订数据造成前视偏差 |
| Financial news and research金融新闻与研究 | Search events, retrieve articles, classify topics搜索事件、获取文章、分类主题 | Publisher, timestamp, URL, language, usage rights发布者、时间戳、URL、语言和使用权 | Duplicate or unlicensed content重复或未授权内容 |
Reference Architecture参考架构
Put MCP between the agent and provider-specific execution把 MCP 放在智能体与供应商执行层之间
The model should see compact tool contracts and structured results—not raw credentials, unbounded schemas, or every provider SDK. 模型应看到精简的工具契约和结构化结果，而不是原始凭据、无限数据结构或所有供应商 SDK。
| Layer层级 | Responsibility职责 | Financial control金融控制 |
|---|---|---|
| Agent host智能体 Host | Plans the task, selects available tools, requests approval where required规划任务、选择可用工具，并在需要时请求批准 | User identity, purpose, workspace policy用户身份、调用目的、工作区策略 |
| MCP clientMCP 客户端 | Negotiates transport, lists tools, sends typed calls, returns results协商传输、列出工具、发送结构化调用并返回结果 | Allowed server list, secret handling, call approval允许的服务器列表、密钥处理和调用审批 |
| Financial data MCP server金融数据 MCP 服务器 | Exposes tools, validates parameters, applies limits, normalizes errors暴露工具、验证参数、执行限制并统一错误 | Scopes, timeouts, row caps, provider routing权限范围、超时、行数上限和供应商路由 |
| Provider and source layer供应商与数据源层 | Executes against SEC, macro, news, fundamentals, or proprietary data调用 SEC、宏观、新闻、基本面或专有数据 | Entitlements, rate limits, attribution, redistribution rules访问权、速率限制、归属与再分发规则 |
| Evidence and audit layer证据与审计层 | Stores request purpose, tool ID, source, execution ID, cost, and outcome保存调用目的、工具 ID、来源、执行 ID、成本和结果 | Traceability, reconciliation, incident review可追踪性、对账与事故复盘 |
MCP supports local stdio and remote Streamable HTTP transports. Transport choice changes deployment and authentication, but it should not change the financial result contract or validation rules.
MCP 支持本地 stdio 与远程 Streamable HTTP 传输。传输方式会改变部署和认证，但不应改变金融结果契约或验证规则。
Tool Contract工具契约
Design results for auditability before adding more tools先为可审计性设计结果，再增加更多工具
A useful tool name and JSON Schema are only the beginning. Financial results need semantic metadata so an agent can distinguish a filing date from a fiscal period, millions from raw units, reported values from estimates, and current values from historical vintages. 清晰的工具名和 JSON Schema 只是起点。金融结果还需要语义元数据，让智能体能够区分提交日期与财务报告期、百万单位与原始单位、实际值与预期值，以及当前值与历史版本。
Return an explicit error code when data is unavailable, delayed, outside entitlement, ambiguous, or too large. Do not convert every provider failure into an empty array; an agent may interpret an empty result as a real absence. 当数据不可用、延迟、超出授权、存在歧义或规模过大时，应返回明确错误码。不要把所有供应商错误都转为空数组，否则智能体可能把空结果误解为事实上的不存在。
Minimum result envelope最小结果封装
{
"data": { "...": "provider result" },
"source": {
"provider": "provider_name",
"source_url": "https://...",
"retrieved_at": "2026-07-30T08:00:00Z"
},
"financial_context": {
"entity_id": "CIK0000320193",
"period": "FY2025",
"currency": "USD",
"units": "USD",
"as_of": "2025-12-31",
"revision": "reported"
},
"execution": {
"tool_id": "resolved_tool_id",
"execution_id": "exec_...",
"status": "success"
}
}
QVeris ProtocolQVeris 协议
Use discover, inspect, probe, and call as separate control points把 discover、inspect、probe 与 call 作为独立控制点
QVeris keeps the MCP surface compact while letting an agent search a larger capability network at runtime. QVeris 保持精简的 MCP 表面，同时允许智能体在运行时搜索更大的能力网络。
01 · discover
Describe the financial capability, not the final parameter values描述金融能力，而不是最终参数值
Search for “retrieve SEC company facts by CIK” or “economic series observations with vintage dates.” Save the returned search_id and candidate tool_id.搜索“按 CIK 获取 SEC 公司事实”或“带历史版本日期的经济序列观测值”，并保存返回的 search_id 与候选 tool_id。
02 · inspect
Review parameters, examples, provider, quality, latency, and billing检查参数、示例、供应商、质量、延迟与计费
Reject a candidate if it lacks the required period, entity identifier, units, or provenance fields. Re-inspect a saved tool before reuse when the task is material.如果候选工具缺少所需报告期、实体标识、单位或来源字段，应拒绝使用。对重要任务复用已保存工具前，应重新 inspect。
03 · probe
Validate parameters and quote before execution执行前验证参数并确认报价
Use probe when a call is expensive, parameters are uncertain, or the workflow must fail before consuming credits. Treat probe success as input validation—not proof that source data is correct.当调用成本较高、参数不确定或工作流必须在消耗积分前失败时使用 probe。probe 成功只代表输入通过验证，不代表源数据一定正确。
04 · call
Execute with the selected tool ID and structured parameters使用选定工具 ID 与结构化参数执行
Preserve the execution ID, provider metadata, billing information, and raw result before transforming values into a memo or model-ready table.在把数据转换为备忘录或模型表格前，保留执行 ID、供应商元数据、计费信息与原始结果。
05 · usage_history / credits_ledger
Reconcile the workflow after the result is used结果使用后完成工作流对账
Use usage history to audit calls and the credits ledger to reconcile final settlement. Link both records to the research run, user, and decision artifact.使用 usage history 审计调用，使用 credits ledger 核对最终结算，并把两类记录关联到研究任务、用户与决策产物。
Before Setup配置前准备
Prepare identity, data scope, and validation before connecting an agent连接智能体前准备好身份、数据范围与验证基准
1
An MCP-compatible client and a valid QVeris API key.兼容 MCP 的客户端与有效的 QVeris API Key。
2
Node.js 18+ for the local @qverisai/mcp package, or remote MCP support for the hosted endpoint.本地 @qverisai/mcp 需要 Node.js 18+；托管端点需要客户端支持远程 MCP。
3
A secret store or environment variable path that keeps API keys out of prompts and source control.使用密钥存储或环境变量，避免 API Key 进入提示词与源代码。
1
Approved data categories, permitted users, and downstream uses.已批准的数据类别、允许用户与下游用途。
2
Golden records or queries for at least three representative workflows.至少三个代表性工作流的基准记录或查询。
3
A policy for delayed data, revisions, missing values, conflicting providers, and human review.针对延迟数据、修订、缺失值、供应商冲突与人工复核的策略。
Setup配置
Connect QVeris MCP locally or through hosted HTTP通过本地进程或托管 HTTP 连接 QVeris MCP
Choose one transport for each environment, keep the key in a secret, reconnect the client, and verify the canonical tools before running financial queries. 为每个环境选择一种传输方式，把密钥保存在安全位置，重新连接客户端，并在执行金融查询前确认标准工具可见。
Use the npm package when the MCP client can start a local process. This works well for desktop clients and development environments.当 MCP 客户端可以启动本地进程时使用 npm 包，适合桌面客户端与开发环境。
{
"mcpServers": {
"qveris": {
"command": "npx",
"args": ["-y", "@qverisai/mcp"],
"env": {
"QVERIS_API_KEY": "YOUR_QVERIS_API_KEY"
}
}
}
}Use the hosted endpoint when the client supports remote MCP and you do not want to manage a local package or background process.当客户端支持远程 MCP，且你不希望管理本地包或后台进程时，使用托管端点。
{
"mcpServers": {
"qveris": {
"type": "http",
"url": "https://mcp.qveris.ai/mcp",
"headers": {
"Authorization": "Bearer YOUR_QVERIS_API_KEY"
}
}
}
}
Verify验证
Reconnect the client and confirm discover, inspect, probe, call, usage_history, and credits_ledger appear. For local clients, QVeris CLI can validate the configuration and probe the MCP process.重新连接客户端，确认可以看到 discover、inspect、probe、call、usage_history 和 credits_ledger。对于本地客户端，可使用 QVeris CLI 验证配置并探测 MCP 进程。
First safe test首次安全测试
Start with discovery and inspection only. Search for a public filing or macroeconomic capability, inspect its parameters and source, then probe before making the first paid or provider-backed call.首次测试只执行发现与检查。先搜索公开监管文件或宏观经济能力，检查参数与来源，再在首次付费或供应商调用前执行 probe。
Worked Example完整示例
Build an earnings research workflow that can be reproduced构建可复现的财报研究工作流
The goal is not merely to generate a summary. The goal is to preserve enough evidence that another analyst or agent can reproduce the result. 目标不只是生成摘要，而是保留足够证据，让另一位分析师或智能体可以复现结果。
Step 1
Resolve the company identity解析公司身份
Map the user’s company name or ticker to stable identifiers such as CIK and exchange-qualified ticker. Stop if the input maps to multiple issuers or share classes.把用户输入的公司名称或股票代码映射为 CIK、交易所限定代码等稳定标识。若输入对应多个发行人或股权类别，应停止并请求确认。
Step 2
Retrieve the latest filing and prior comparable period获取最新文件与上一可比期间
Record form type, accession number, filing timestamp, fiscal period, and source URL. Compare the same form and fiscal period rather than whichever document is newest.记录表格类型、accession 编号、提交时间、财务报告期和来源 URL。比较相同表格和财务期间，而不是简单选择最新文档。
Step 3
Fetch normalized facts with units and taxonomy获取带单位与分类标准的规范化事实
Retrieve revenue, operating income, cash flow, and other required facts. Preserve the original taxonomy concept, units, currency, start date, end date, and whether the value was restated.获取收入、营业利润、现金流等所需事实，并保留原始分类概念、单位、币种、开始日期、结束日期以及是否重述。
Step 4
Add earnings and macro context without mixing clocks加入财报与宏观背景，但不要混用时间口径
Keep event time, publication time, reporting period, and macro vintage separate. A revised macro series observed today must not be used as if it were available on the historical decision date.分别保留事件时间、发布时间、财务报告期与宏观数据版本。今天看到的修订后宏观序列，不能假装在历史决策日期已经可用。
Step 5
Generate conclusions with citations and uncertainty生成带引用与不确定性的结论
Every material claim should point to a result record or source URL. Mark estimates, missing values, provider conflicts, and transformations instead of smoothing them into confident prose.每个重要结论都应指向结果记录或来源 URL。对预估值、缺失值、供应商冲突和数据转换进行标记，不要把它们润色成过度确定的文字。
Step 6
Store the run manifest保存运行清单
Persist user purpose, identifiers, tool IDs, parameters, execution IDs, retrieved timestamps, source links, transformations, validation outcomes, and final settlement. This manifest is the audit trail for the memo.持久化用户目的、标识符、工具 ID、参数、执行 ID、获取时间、来源链接、转换、验证结果与最终结算。该清单就是研究备忘录的审计轨迹。
Validation Scorecard验证评分表
Test financial meaning, not only HTTP success验证金融含义，而不只是 HTTP 成功
A 200 response can still contain the wrong entity, period, unit, revision, or licensing context. Use pass-or-fail evidence before production. 即使返回 200，也可能包含错误实体、期间、单位、版本或授权背景。上线前应使用明确的通过/失败证据。
| Check检查项 | Pass evidence通过证据 | Failure action失败处理 |
|---|---|---|
| Entity identity实体身份 | Ticker, legal name, exchange, and stable identifier agree股票代码、法定名称、交易所与稳定标识一致 | Request disambiguation; do not guess请求消歧，不做猜测 |
| Period and timestamp期间与时间戳 | Fiscal period, filing date, event time, and retrieval time are separate财务期间、提交日期、事件时间和获取时间分别记录 | Block comparison until clocks align时间口径一致前阻止比较 |
| Units and currency单位与币种 | Raw units are preserved and transformations are explicit保留原始单位并明确记录转换 | Return a typed unit error返回明确的单位错误 |
| Source provenance来源追踪 | Provider, source URL, tool ID, and execution ID are present供应商、来源 URL、工具 ID 和执行 ID 齐全 | Do not publish unsupported claims不得发布无来源支持的结论 |
| Golden-query parity基准查询一致性 | Representative totals match trusted records within defined tolerance代表性总量在规定容差内与可信记录一致 | Quarantine the tool or provider route隔离该工具或供应商路由 |
| Error behavior错误行为 | Missing, delayed, unauthorized, ambiguous, and rate-limited cases are distinct缺失、延迟、未授权、歧义和限流均有不同错误 | Fail closed; never invent a value失败关闭，绝不编造数值 |
| Audit reconciliation审计对账 | Usage history, credits, and research manifest reconcile调用历史、积分记录与研究清单一致 | Hold downstream delivery for review暂停下游交付并复核 |
Production Controls生产控制
Treat data rights, access, and point-in-time integrity as first-class controls把数据权利、访问控制与时点完整性作为一级控制
Keep provider and QVeris keys outside prompts. Bind access to the user, workspace, purpose, and allowed capability class. Rotate compromised keys and start a new MCP session after a key change.让供应商与 QVeris 密钥远离提示词。把访问绑定到用户、工作区、用途和允许的能力类别。密钥泄露后应轮换，并在更换密钥后创建新的 MCP 会话。
MCP does not grant rights to redistribute provider data, full articles, transcripts, or exchange feeds. Enforce the underlying provider’s display, retention, attribution, and user-entitlement terms.MCP 不会授予再分发供应商数据、全文文章、电话会记录或交易所行情的权利。必须执行底层供应商关于展示、保留、归属与用户授权的条款。
Store the version or vintage available on the decision date. This matters for revised macro series, restated financials, corrected filings, late transcripts, and backtests.保存决策日期当时可用的数据版本。这对修订后的宏观序列、重述财务数据、更正文件、延迟电话会记录和回测非常重要。
Apply provider-aware caching, pagination, backoff, and bulk-download strategies. For SEC automated access, identify the client and follow current fair-access guidance rather than scraping aggressively.根据供应商设置缓存、分页、退避与批量下载策略。对 SEC 自动访问，应识别客户端并遵守当前公平访问指引，而不是高频抓取。
Require review before a result triggers a trade, client communication, valuation change, compliance filing, or material capital decision. Tool availability is not authorization to act.当结果将触发交易、客户沟通、估值变更、合规申报或重大资本决策时，应要求人工复核。工具可用不等于获准执行行动。
If the source is unavailable or ambiguous, return the typed failure and next action. Never let the model fill a missing quarter, currency, estimate, or price from memory.如果来源不可用或存在歧义，应返回明确失败类型与下一步操作。绝不能让模型凭记忆补全缺失季度、币种、预估值或价格。
Failure Modes失败模式
Avoid the mistakes that make finance agents confidently wrong避免让金融智能体“自信地犯错”
Most production failures are not caused by MCP transport. They come from ambiguous entities, incomplete result contracts, ignored licensing, mixed time periods, silent provider errors, and summaries that lose source evidence. 大多数生产故障不是由 MCP 传输造成的，而是来自实体歧义、结果契约不完整、忽视许可、混用时间期间、供应商错误被静默处理，以及摘要丢失来源证据。
1
One giant finance tool: dozens of optional parameters create ambiguous calls. Split by operation and keep schemas narrow.一个巨型金融工具：大量可选参数导致调用歧义。应按操作拆分并保持结构精简。
2
Ticker-only identity: tickers change and can be reused. Resolve stable identifiers first.只使用股票代码：股票代码会变化或被复用，应先解析稳定标识。
3
Empty array on provider failure: distinguish “no record” from timeout, entitlement, and rate-limit errors.供应商失败时返回空数组：应区分“没有记录”、超时、未授权和限流。
4
No raw evidence: a polished memo without source URLs and execution IDs cannot be audited.没有原始证据：缺少来源 URL 与执行 ID 的精美报告无法审计。
5
Provider fallback without equivalence: do not switch sources unless units, timing, adjustments, and license are compatible.未验证等价性就回退供应商：单位、时间、调整方式与许可兼容前，不得切换来源。
Build or Buy自建或采购
Choose the smallest architecture that preserves your required controls选择能够保留所需控制的最小架构
QVeris is useful when provider-backed capabilities must be discovered across categories. It is not automatically the right answer for every private dataset or ultra-low-latency workload. 当智能体需要跨类别发现供应商支持的能力时，QVeris 很有价值；但它并不自动适合所有私有数据集或超低延迟工作负载。
| Option方案 | Best fit最适合 | Advantages优势 | Trade-offs取舍 |
|---|---|---|---|
| Direct provider API直接供应商 API | One known dataset and a stable application workflow一个已知数据集与稳定应用流程 | Full endpoint control, direct vendor support, minimal routing完整端点控制、直接供应商支持、最少路由 | Custom adapters, separate auth, billing, errors, and monitoring需要自建适配器，并分别处理认证、计费、错误与监控 |
| Custom financial MCP server自建金融 MCP 服务器 | Private data, proprietary models, fixed internal semantics私有数据、专有模型与固定内部语义 | Maximum policy and schema control inside your infrastructure在自有基础设施内获得最大策略与结构控制 | You own protocol upgrades, provider integrations, testing, security, and operations自行承担协议升级、供应商集成、测试、安全与运维 |
| QVeris MCP serverQVeris MCP 服务器 | Agents that discover and call multiple provider-backed capabilities需要发现并调用多个供应商能力的智能体 | Compact MCP surface, runtime discovery, inspection, probe, call, usage audit, and credits ledger精简 MCP 表面、运行时发现、检查、探测、调用、使用审计与积分账本 | Private semantics still need your own layer; provider terms and validation still apply私有语义仍需自建层；供应商条款与验证要求仍然适用 |
| Hybrid混合方案 | Private research data plus external financial capabilities私有研究数据加外部金融能力 | Keep internal truth systems private while using QVeris for external capability coverage保持内部事实系统私有，同时使用 QVeris 扩展外部能力 | Requires clear routing, identity, provenance, and conflict rules需要清晰的路由、身份、来源与冲突处理规则 |
FAQ
Financial data MCP server questions金融数据 MCP 服务器常见问题
What is a financial data MCP server?什么是金融数据 MCP 服务器？
It is an MCP server that exposes financial-data operations as discoverable, schema-defined tools for AI applications. Typical operations include resolving companies, retrieving filings and facts, reading earnings events, fetching macro series, and searching financial news.
它是一种 MCP 服务器，把金融数据操作暴露为 AI 应用可发现、具有参数结构的工具。典型操作包括解析公司、获取监管文件与财务事实、读取财报事件、获取宏观序列和搜索金融新闻。
What financial data should an MCP server expose?金融数据 MCP 服务器应暴露哪些数据？
Start with narrow tools for entity resolution, company fundamentals, filings, earnings, macroeconomic series, and news. Each result should include source, timestamp, entity, period, units, currency, revision state, and execution metadata. Treat low-latency market data as a separate workload.
应从实体解析、公司基本面、监管文件、财报、宏观经济序列和新闻等范围明确的工具开始。每个结果都应包含来源、时间戳、实体、报告期、单位、币种、修订状态和执行元数据。低延迟市场数据应作为单独工作负载处理。
How does QVeris provide financial data through MCP?QVeris 如何通过 MCP 提供金融数据？
QVeris exposes discover, inspect, probe, and call through its MCP server. The agent searches for a capability, reviews its schema and quality signals, validates parameters and price when needed, executes the selected tool, then reviews usage history and credit settlement.
QVeris 通过 MCP 服务器暴露 discover、inspect、probe 和 call。智能体先搜索能力，检查参数结构与质量信号，在需要时验证参数和价格，执行选定工具，最后检查调用历史与积分结算。
Is a financial data MCP server the same as a market data MCP server?金融数据 MCP 服务器等同于市场数据 MCP 服务器吗？
No. Financial data includes filings, fundamentals, earnings, macro data, news, and research context. Market data focuses on quotes, trades, bars, order books, exchange sessions, streaming, and latency. The systems can share an MCP boundary but need different freshness, licensing, and validation controls.
不完全相同。金融数据包括监管文件、基本面、财报、宏观数据、新闻和研究上下文；市场数据聚焦报价、成交、K 线、订单簿、交易时段、流式传输和延迟。两者可以共享 MCP 边界，但时效、许可和验证控制不同。
Can an AI agent trust financial data returned through MCP?AI Agent 可以直接信任 MCP 返回的金融数据吗？
Not without validation. MCP standardizes access, but correctness depends on the provider, identifiers, units, periods, revisions, and transformations. Production workflows should compare representative outputs with trusted records and preserve source and execution evidence.
不能在未验证的情况下直接信任。MCP 规范访问方式，但正确性仍取决于供应商、标识符、单位、期间、修订与转换。生产工作流应把代表性输出与可信记录比较，并保留来源和执行证据。
Do I still need provider licenses and access controls?仍然需要供应商许可和访问控制吗？
Yes. MCP does not override provider terms, redistribution rights, exchange rules, user entitlements, retention limits, or attribution requirements. Apply these controls to every underlying source and downstream use.
需要。MCP 不会覆盖供应商条款、再分发权、交易所规则、用户授权、保留限制或归属要求。必须对每个底层来源和下游用途执行这些控制。
When should I use QVeris instead of building a custom finance MCP server?什么时候应使用 QVeris，而不是自建金融 MCP 服务器？
Use QVeris when agents need runtime discovery across multiple provider-backed capabilities and you want one maintained MCP, REST, SDK, and CLI protocol. Build a custom server when proprietary datasets, private calculations, fixed internal schemas, or fully isolated controls are the core requirement. A hybrid model often works best.
当智能体需要在多个供应商能力之间运行时发现，并希望使用统一维护的 MCP、REST、SDK 与 CLI 协议时，适合使用 QVeris。当核心需求是专有数据、私有计算、固定内部结构或完全隔离的控制时，适合自建服务器。很多团队最终采用混合模式。
Sources and Next Steps资料与下一步
Verify the protocol, source rules, and implementation details核对协议、数据源规则与实施细节
Model Context Protocol architectureModel Context Protocol 架构
MCP server concepts: tools, resources, and promptsMCP 服务器概念：tools、resources 与 prompts
SEC EDGAR APIs and XBRL dataSEC EDGAR API 与 XBRL 数据
QVeris MCP Server documentationQVeris MCP 服务器文档
Hosted MCP endpoint and client configuration托管 MCP 端点与客户端配置
How to use QVeris MCP Server如何使用 QVeris MCP 服务器
SEC filing API for AI agents面向 AI Agent 的 SEC 文件 API
Best financial data APIs for AI agents面向 AI Agent 的金融数据 API 指南
Start with Evidence从证据开始
Connect one financial workflow, then prove it先连接一个金融工作流，再验证它
Configure QVeris MCP, discover one filing or fundamental capability, inspect and probe it, run a golden query, and preserve the complete execution record before expanding to more providers. 配置 QVeris MCP，发现一个监管文件或基本面能力，完成 inspect 与 probe，运行一条基准查询，并在扩展更多供应商前保留完整执行记录。
Discover
Find the right provider-backed capability找到合适的供应商能力
Inspect + Probe
Check schema, quality, parameters, and quote检查结构、质量、参数与报价
Call + Audit
Execute, validate, and reconcile the result执行、验证并核对结果