Signaliz

io.github.signaliz由 GitHub 账号 signaliz 发布v2.0.1更新于 Oct 25, 2026

B2B lead generation, email verification, company enrichment, and agentic GTM Ops.

已验证Streamable HTTP可网页运行效率与工作流数据与分析商业与电商

概览

AI 生成的概览

Signaliz 是托管的 B2B 数据平台，其 MCP 服务器让助手查找并验证潜在客户、丰富公司信息并运行 GTM 自动化例程。

可以做什么
查找并验证邮箱、联系人和公司信息，支持最多 5,000 条记录的批量任务，可通过轮询或 webhook 获取结果。发现 B2B 专业人士、目标客户和本地商家，用于构建外联名单。上传名单，创建、运行并查看已保存的自动化系统和流水线运行。创建、串联、触发并流式获取目标驱动的 Ops 例程，并发出可唤醒例程的事件。管理审批、条目反馈、应用连接和连接器操作。管理 GTM 活动手册、ICP、预算、屏蔽名单和持久化代理记忆。
适用场景
适合 GTM 和销售运营场景：让助手在 Signaliz 工作区内构建经过验证的潜在客户名单、调研公司，或运行周期性的数据丰富与外联准备流程。也适合需要带审批和额度预算的代理驱动自动化的团队。
运行要求
由 Signaliz 托管的远程 Streamable HTTP 端点。认证使用 Signaliz API 密钥，以 Authorization Bearer 令牌发送，可在 Settings > Developer > API Access 中创建，或在客户端支持时使用 OAuth。可选的 npm stdio 桥接需要 Node.js 和 SIGNALIZ_API_KEY 环境变量。需要带 GTM 额度的 Signaliz 套餐。
安装前请注意
按量计费的供应商支持操作（如邮箱验证、发现、数据丰富和自定义 AI 提示）会消耗 GTM 额度，额外额度可购买。许多工具会写入或更改数据：上传名单、创建和运行系统、连接应用、执行应用操作、更新屏蔽名单、审批工作和删除例程。API 密钥可访问工作区，须妥善保密。数据会发送至 Signaliz 及已连接的第三方应用。

安装

在 SourceWeft 中

  1. 打开 控制台中的 Signaliz，将其添加到工作区。
  2. 连接 所需的 凭据。它们会被私密存储，绝不会在公开页面上展示。
  3. 为需要使用其工具的对话启用该服务。

Web executable，通过 Streamable HTTP。 远程服务在工作区中配置后即可从网页运行时运行。

交给智能体安装

把这段提示词粘贴到 Claude Code、Codex、Cursor 等智能体，或 SourceWeft 对话中。智能体会读取该服务器的安装指南，先向你展示发布者、端点和登录方式，经你同意后再把它添加到你的客户端。密钥不会写进任何命令。

阅读 https://sourceweft.com/mcp/io.github.signaliz%2Fsignaliz-mcp/install.md，并按其中说明把这个 MCP 服务器接入我的客户端。修改任何配置之前，先告诉我它的发布者、端点和登录方式，等我确认。不要把 API 密钥写进命令或我会提交的文件；问我密钥应从哪里获取。

查看智能体所依据的安装指南

其他 MCP 客户端

把它添加到你客户端的 mcpServers 配置中。 客户端会在首次连接时完成 OAuth 登录。

{
  "mcpServers": {
    "signaliz": {
      "type": "http",
      "url": "https://api.signaliz.com/functions/v1/signaliz-mcp"
    }
  }
}

README

Signaliz MCP

Signaliz is a B2B data quality, enrichment, and agentic GTM platform for go-to-market teams and AI agents. The Signaliz MCP server gives MCP clients access to 60+ Signaliz API capabilities for verified lead generation, email verification, company enrichment, data-quality pipelines, Ops routines, approvals, ICPs, app actions, and workspace controls.

This repository is the public distribution manifest for Signaliz MCP. It contains registry metadata, indexable tool schemas, and setup instructions for MCP directories and clients. The production server is hosted by Signaliz; this repository does not contain the private server implementation.

Hosted Endpoint

text
https://api.signaliz.com/functions/v1/signaliz-mcp

Transport: Streamable HTTP

Authentication: Signaliz API key or OAuth, depending on client support.

Quick Start

Create an API key in Signaliz from Settings > Developer > API Access, then connect your MCP client to:

text
https://api.signaliz.com/functions/v1/signaliz-mcp?api_key=YOUR_API_KEY

Clients that support headers can connect to the base endpoint and send:

text
Authorization: Bearer YOUR_API_KEY

Clients that support OAuth discovery can use the protected-resource metadata advertised by the server.

Local Stdio Bridge

For clients that prefer local stdio MCP servers, install the npm bridge:

bash
claude mcp add signaliz -e SIGNALIZ_API_KEY=sk_your_key -- npx -y @signaliz/mcp-server

Claude Desktop, Cursor, Windsurf, Cline, and other stdio clients can use:

json
{  "mcpServers": {    "signaliz": {      "command": "npx",      "args": ["-y", "@signaliz/mcp-server"],      "env": {        "SIGNALIZ_API_KEY": "sk_your_key"      }    }  }}

What Signaliz Does

Signaliz exposes tools for:

  • Finding verified emails for known contacts.
  • Verifying deliverability for one email or thousands of emails.
  • Discovering buying signals such as hiring, funding, product launches, partnerships, leadership changes, expansion, acquisitions, awards, regulatory events, and earnings.
  • Finding B2B professionals, target accounts, local businesses, or outreach-ready leads.
  • Running custom AI prompts across records with structured output fields.
  • Calling external APIs through generic HTTP tools.
  • Uploading lists and running governed Signaliz systems against them.
  • Creating autonomous Ops Routines that work on a GTM goal over time.
  • Chaining routines, streaming results, emitting events, and resolving approvals.
  • Connecting apps, discovering available actions, and executing app actions through Signaliz.
  • Managing budgets, blocklists, ICPs, campaign books, and durable agent memory.

Every tool accepts output_format, which can be markdown or json. Use json when another agent, script, or workflow needs structured output.

Access Model

Signaliz plans include full API and MCP access, unlimited workspaces, unlimited integrations, audit trail and MCP analytics, BYOK OpenRouter support, and all core platform features. Plans differ by included monthly GTM credits, and additional credits can be purchased for metered provider-backed actions such as email verification, person or company discovery, company signal enrichment, monitoring, custom AI prompts, and Lead Finder.

Signaliz is not positioned as unlimited-volume or all-you-can-use data. The product is designed around accuracy, timing, governed workflows, and transparent GTM credit usage.

Capability Index

The full indexable tool schema lives in tools/tools.json. It is generated from the actual @signaliz/[email protected] tools/list response so directories can index the same names, descriptions, and input schemas exposed by the npm bridge.

Email, Contact, and Company Enrichment

ToolWhat it does
find_contacts_with_emailFinds contacts at a company and returns verified work emails.
find_emails_with_verificationFinds and verifies one professional email address from a name, company domain, or LinkedIn URL.
verify_emailVerifies one email address for deliverability.
enrich_company_signalsResearches a company for structured signals and source-backed developments.
company_intelligenceRuns AI-powered company research from websites, domains, LinkedIn URLs, and extra URLs.
execute_primitiveRuns small synchronous jobs for email verification, email finding, signal enrichment, custom AI prompts, data cleaning, or HTTP requests.
generic_http_requestExecutes one custom HTTP request with headers, auth, body templates, variables, timeout, and response extraction.
custom_ai_promptRuns a custom AI prompt over records with model selection, structured output, and optional Signaliz AI Search context.

Batch Jobs

ToolWhat it does
find_and_verify_emailsFinds and verifies emails for up to 5,000 contacts asynchronously.
verify_emailsVerifies up to 5,000 emails asynchronously.
enrich_company_signals_batchEnriches signals for up to 5,000 companies asynchronously.
batch_http_requestExecutes up to 5,000 HTTP requests asynchronously.
check_job_statusPolls async jobs, reports progress, and retrieves paginated results.

Batch tools return a job_id immediately. Poll with check_job_status, or pass callback_url and optional callback_secret on supported batch tools to receive a completion webhook.

Lead and Account Discovery

ToolWhat it does
generate_leadsFinds B2B professionals with company data and verified emails for outreach-ready lead lists.
generate_local_leadsFinds local businesses and verified contact emails from Google Maps-style local search.
find_people_blitzFinds professional profiles without verified emails for research and list building.
find_companies_blitzFinds company records for account discovery, market sizing, and TAM research.

Lists, Systems, and Pipeline Runs

ToolWhat it does
list_capabilitiesLists available Signaliz capabilities, schemas, categories, and credit costs.
upload_dataUploads inline records, public CSV URLs, Google Sheets URLs, or Google Drive share URLs as workspace lists.
list_systemsLists saved Signaliz automation systems in the workspace.
get_systemGets a saved system's phases, required inputs, and configuration.
create_systemCreates a saved automation system from a plain-English description.
run_systemRuns a saved system using inline data, a saved list, the latest workspace list, or prior run data.
get_runGets one system run's status and results.
list_runsLists recent system runs with status, timing, and summaries.

Ops Routines as Agents

ToolWhat it does
create_routineCreates a goal-driven Ops Routine with cadence, policy guardrails, wake events, and output sinks.
list_routinesLists Ops Routines by status.
get_routineGets full routine details, policy, sinks, recent ticks, and outcomes.
update_routineUpdates routine name, goal, cadence, status, policy, sinks, or wake events.
run_routine_nowTriggers an Ops Routine immediately with an optional one-off instruction.
delete_routineSoft-archives or permanently deletes a routine after confirmation.
get_routine_ticksLists historical routine ticks with status, record counts, credits, and errors.
get_tick_itemsFetches records produced by a specific tick.
get_last_tick_itemsFetches records from the latest successful tick.
get_routine_resultsAgent-friendly helper that combines routine metadata with latest produced items.
agent_workflowCreates, activates, triggers, optionally waits for, and optionally cleans up a one-shot Ops Routine.
stream_routine_resultsReturns an SSE endpoint for real-time item events from a running routine tick.
chain_routinesChains two routines so upstream output feeds downstream seed items.
get_chain_statusShows chained routine execution status and per-stage item counts.
emit_eventEmits an event that can wake routines subscribed through wake_on_events.

Approvals and Feedback

ToolWhat it does
approvals_listLists pending, approved, or rejected human-in-the-loop approvals.
approvals_decideApproves or rejects one or more pending approval tokens.
set_item_feedbackStores feedback on a produced lead or item so future routine ticks can improve.

App Connections and Actions

ToolWhat it does
list_connectionsLists active app connections in the workspace.
list_available_connectorsLists available connectors across CRM, marketing, communication, productivity, analytics, development, support, finance, HR, ecommerce, email, and sales.
list_connectorsLegacy alias for available connector listing.
connect_appStarts the app connection flow and returns authentication instructions.
discover_actionsLists available entities and actions for a connected app.
execute_connection_actionExecutes read or write actions against a connected app.
list_connection_audit_logShows connection lifecycle events, tests, and errors for debugging.

GTM Books and ICPs

ToolWhat it does
quickstart_gtm_bookLaunches a coordinated GTM Campaign Book made of multiple Ops Routines.
list_booksLists GTM Campaign Books in the workspace.
book_roll_upGets aggregate performance metrics across a campaign book.
list_icpsLists Ideal Customer Profiles and optional Octave import context.
get_icpGets full ICP details, criteria, pain points, triggers, titles, departments, geography, and related playbook data.
import_icps_from_octaveImports Octave playbooks as Signaliz ICPs.

Workspace Controls

ToolWhat it does
budget_get_statusGets current credit balance, spend rate, limits, and projected runway.
budget_set_limitSets daily, monthly, or system-specific credit limits and alert thresholds.
manage_blocklistAdds, removes, or lists suppressed domains and emails.
write_agent_memoryStores durable agent memory for future sessions.
query_agent_memorySearches stored agent memories with semantic retrieval.
list_output_sinksDeprecated compatibility tool; Signaliz now returns JSON, CSV artifacts, and webhooks.

Registry Metadata

Safety Notes

Read-only enrichment and discovery tools do not mutate your external systems. Tools that upload data, create or run systems, connect apps, execute app actions, update blocklists, approve work, or delete routines can change Signaliz workspace resources or connected systems and should be used with your normal approval controls.

Links

来源：README.md，提交 75bc98e

工具

0
该服务器需要登录，因此 SourceWeft 无法在没有账号的情况下读取其工具，也未收录。已连接账号所能看到的内容可能因账号及其权限而异。

版本历史

1
  1. v2.0.1最新Sep 16, 2026

评分

已满足适用检查项的 75%

该评分衡量此收录条目的完整程度，依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价，也不表示服务器安全或当前可用：可连通性是带日期的检查结果，评分也不包含安装量或用户评价。

  • 近 30 天内可连通必需25 分
  • 已声明认证方式必需10 分
  • 已列出工具15 分

    请让服务器在无凭据时响应 tools/list，或发布可在沙箱中无凭据启动的软件包。

  • 描述必需15 分
  • 图标10 分
  • 文档15 分
  • 许可证10 分

    请在仓库或注册中心条目中声明许可证。