Introduction
Spectre is a hosted MCP server that gives AI agents direct access to LinkedIn people, company, and job data — no scraping setup, no LinkedIn cookies, no provider accounts required.
What Spectre does
Spectre exposes 9 tools across three resource types. Your agent can look up people profiles, search for candidates or leads, retrieve company information, and pull job posting details — all through a single MCP connection.
People
Look up a LinkedIn profile by username, search by name, or run structured searches by title, company, and location. Four tools give you standard and raw access at every level.
Companies
Retrieve a company's full LinkedIn profile — name, industry, size, headquarters, website, and more. Standard and raw variants available.
Jobs
Search jobs by keyword and location, or get full details for a specific posting including pay range, description, and applicant count. Standard and raw variants plus a search tool.
How it works
Spectre runs as a hosted MCP server. Your agent connects via the Model Context Protocol — the same way it connects to filesystems, databases, and APIs. Once connected, your agent can call Spectre tools just like any other MCP tool. No extension, no plugin, no separate service to install.
Authentication is handled through OAuth (for Claude web and mobile) or an API key (for desktop and CLI clients). Both paths give your agent access to the same 9 tools with identical behavior.
Use cases
Sales research
Enrich leads with real LinkedIn data before outreach. Pull company context and decision-maker profiles in a single agent workflow.
Recruiting
Search for candidates by title and location, then drill into full profiles. Surface job postings to understand market demand.
Market intelligence
Research companies at scale, track hiring signals, and monitor industry movement — all through your agent.