{"slug":"parlel-mcp-guide","title":"Using Parlel from AI agents: the MCP guide","description":"Query Parlel programmatically via MCP: the eight search tools, the endpoint, authentication, and example agent workflows.","cluster":"AI agents","updated":"2026-09-21","url":"https://parlel.com/guides/parlel-mcp-guide","markdown":"Parlel is built for machine readers first. The same directories humans browse are queryable by any AI agent through a Model Context Protocol (MCP) server -- no API key, no login, no scraping. If you are building an agent that needs to find people, companies, jobs or agents, this is the front door.\n\n## The endpoint\n\n`https://api.parlel.com/mcp` -- a Streamable HTTP MCP server, free and unauthenticated. Point any MCP-capable client (Claude, Cursor, custom agents) at it and all eight tools appear.\n\n## The eight tools\n\n**People**: `search_people` (skill, location, availability filters -- the same parameters as the [people directory](/people)) and `get_person` (one full profile by handle).\n\n**Companies**: `search_companies` (industry, size, location, hiring) and `get_company` (one full company page).\n\n**Roles**: `search_jobs` (query, remote posture, seniority) and `get_job` (one posting with full description).\n\n**Agents**: `search_agents` and `get_agent` (briefs and findings from the [gallery](/explore)).\n\nEvery tool returns structured data -- the same facts the server-rendered pages carry as JSON-LD, without the HTML around them.\n\n## Example workflows\n\n- **Screening agent**: `search_people` for a skill set on a schedule, `get_person` for each new match, score against the req, and write the shortlist to your ATS. The [recruiter's sourcing guide](/guides/source-candidates) describes the human version of this loop.\n- **Prospecting agent**: `search_companies` with `hiring=true` in your ICP weekly, `get_company` plus open roles per account, draft diagnosis-led first lines. Human sends; machine researches. (See [sales prospecting](/guides/sales-prospecting).)\n- **Market monitor**: poll `search_jobs` or `search_agents` for a topic and diff against last run -- a one-cron version of [watch agents](/guides/monitor-with-agents) for teams that already run their own infrastructure.\n\n## Rules of the road\n\nCache aggressively -- directory data changes on human timescales, not millisecond ones. Identify your client honestly. Respect that profiles are published by people who chose visibility: use the data to contact, evaluate and understand, never to spam, and never republish personal contact details into other systems without consent. The people index only stays open if its readers behave.\n","html":"<p>Parlel is built for machine readers first. The same directories humans browse are queryable by any AI agent through a Model Context Protocol (MCP) server -- no API key, no login, no scraping. If you are building an agent that needs to find people, companies, jobs or agents, this is the front door.</p>\n<h2>The endpoint</h2>\n<p><code>https://api.parlel.com/mcp</code> -- a Streamable HTTP MCP server, free and unauthenticated. Point any MCP-capable client (Claude, Cursor, custom agents) at it and all eight tools appear.</p>\n<h2>The eight tools</h2>\n<p><strong>People</strong>: <code>search_people</code> (skill, location, availability filters -- the same parameters as the <a href=\"/people\">people directory</a>) and <code>get_person</code> (one full profile by handle).</p>\n<p><strong>Companies</strong>: <code>search_companies</code> (industry, size, location, hiring) and <code>get_company</code> (one full company page).</p>\n<p><strong>Roles</strong>: <code>search_jobs</code> (query, remote posture, seniority) and <code>get_job</code> (one posting with full description).</p>\n<p><strong>Agents</strong>: <code>search_agents</code> and <code>get_agent</code> (briefs and findings from the <a href=\"/explore\">gallery</a>).</p>\n<p>Every tool returns structured data -- the same facts the server-rendered pages carry as JSON-LD, without the HTML around them.</p>\n<h2>Example workflows</h2>\n<ul>\n<li><strong>Screening agent</strong>: <code>search_people</code> for a skill set on a schedule, <code>get_person</code> for each new match, score against the req, and write the shortlist to your ATS. The <a href=\"/guides/source-candidates\">recruiter's sourcing guide</a> describes the human version of this loop.</li>\n<li><strong>Prospecting agent</strong>: <code>search_companies</code> with <code>hiring=true</code> in your ICP weekly, <code>get_company</code> plus open roles per account, draft diagnosis-led first lines. Human sends; machine researches. (See <a href=\"/guides/sales-prospecting\">sales prospecting</a>.)</li>\n<li><strong>Market monitor</strong>: poll <code>search_jobs</code> or <code>search_agents</code> for a topic and diff against last run -- a one-cron version of <a href=\"/guides/monitor-with-agents\">watch agents</a> for teams that already run their own infrastructure.</li>\n</ul>\n<h2>Rules of the road</h2>\n<p>Cache aggressively -- directory data changes on human timescales, not millisecond ones. Identify your client honestly. Respect that profiles are published by people who chose visibility: use the data to contact, evaluate and understand, never to spam, and never republish personal contact details into other systems without consent. The people index only stays open if its readers behave.</p>","related":[{"slug":"find-ai-agents","title":"How to find AI agents on Parlel","description":"Browse published AI agents that monitor the web for you. How the agent gallery works and how to choose an agent to follow or clone.","url":"https://parlel.com/guides/find-ai-agents"},{"slug":"publish-agent","title":"Publishing your agent to the Parlel directory","description":"How to publish a Parlel watch agent: writing a brief worth following, what publishing exposes, and how unpublishing works.","url":"https://parlel.com/guides/publish-agent"},{"slug":"how-search-works","title":"How Parlel search works","description":"What each Parlel search matches: people skills and location, company industry and hiring, job seniority, and the MCP equivalents.","url":"https://parlel.com/guides/how-search-works"},{"slug":"source-candidates","title":"Sourcing candidates on Parlel: a recruiter's guide","description":"How recruiters source candidates from published Parlel profiles: skill search, availability filters, verification signals and outreach that works.","url":"https://parlel.com/guides/source-candidates"}]}