Instinct MCP · Documentation
Using the Instinct MCP
The Instinct MCP gives your AI assistant read-only access to Instinct's managed-services (MSP) market intelligence — company profiles, maturity scores, technographics, reputation, and market sizing — over a simple credit-metered API. This page covers what it is, how to connect, how credits work, and every tool you can call.
Overview
What the Instinct MCP is
A read-only window into Instinct's MSP intelligence, priced by the credit.
Instinct continuously researches Managed Service Providers across the English-speaking market and builds a deep, current profile of each one. The MCP exposes that research to any MCP-compatible AI assistant (Claude, ChatGPT, Gemini, Grok, Cursor, and others) as a set of read-only tools. You ask your assistant a question; it calls the right tool and answers from Instinct's data.
It's company-first — built to help you understand accounts and territories, and it returns company-level business data only (no personal contact information). Every call draws from your credit balance, where 1 credit = $0.01. Cheaper tools are quick lookups; the richer analytic pulls cost more.
Read-only, always
Every tool only reads — the MCP never changes anything. You can't break anything by exploring.
Setup
Connect a client
Point any MCP client at the endpoint and sign in — no API keys to copy.
https://mcp.mspmcp.com/mcpSettings → Connectors → Add custom connector, paste the endpoint, and sign in. Or add it in Claude Code with claude mcp add --transport http instinct https://mcp.mspmcp.com/mcp.
Enable Developer Mode → Connectors → Create, choose “MCP (remote server),” paste the endpoint, and sign in.
In Gemini, open Settings → Connected apps → Custom apps, paste the endpoint, and authorize.
In Grok, open Connectors → New connector → Custom, paste the endpoint, and connect.
On first connect your client opens an Instinct sign-in and consent screen. Approve access — that links the client to your account and its credits. No keys to paste. Don't have an account yet? Sign up.
Billing
Credits & pricing
Monthly credits that reset each cycle — spend them on the tools you call.
Instinct MCP runs on credits: 1 credit = $0.01. Each tool has a fixed credit cost (listed below), and a few are priced per result returned. Your plan grants a monthly allotment of credits that resets each billing cycle (use-or-lose).
See the current plans and monthly credit allotments on the Pricing page.
Reference
The tools
Everything you can ask for, grouped by what it does — with each tool's credit cost and an example question.
Company lookups & profiles
Find a company, then pull the enriched firmographic and service detail behind it.
get_company2 creditsFinds one company by domain or ID; returns its name, location, and whether it's an MSP. The cheap first step before a deeper pull.
Ask: “Look up simpsoncomputing.com.”
get_company_profile_full250 creditsThe full enriched profile: description, focus, target market, size, locations, socials, and the composite FCML band.
Ask: “Give me the full profile for this company.”
get_company_services100 creditsThe company's normalized MSP services, each with a category — cleaner than raw technology text.
Ask: “What services does this MSP offer?”
get_detected_vendors150 creditsThe third-party technology vendors detected (RMM, PSA, EDR, backup…), with detection method and confidence.
Ask: “Which RMM and PSA tools does this MSP use?”
get_tech_profile150 creditsThe detected technographic profile — the tools and services found on the company.
Ask: “What's this company's tech stack?”
Scores & benchmarks
The Instinct analytic layer: how established a company looks, and how it compares.
get_fcml_score_breakdown400 creditsThe FCML score: the composite band plus every per-signal sub-score. The core analytic layer.
Ask: “Break down this company's FCML score.”
get_company_benchmark150 creditsHow a company ranks on one signal versus a cohort — its band plus the cohort percentile. The cohort auto-widens if it's too small.
Ask: “How does this company's LinkedIn presence compare to its peers?”
get_territory_health100 creditsTerritory Health (THS): banded local-market strength with the explainability behind it (radius, ZIP count, population, urbanicity).
Ask: “How strong is this MSP's local market?”
Reputation, social & hiring
How a company shows up publicly — reviews, social reach, and hiring activity.
get_reputation125 creditsCross-source reputation: Google Maps, Glassdoor, and Indeed ratings + review counts.
Ask: “What's this company's online reputation?”
get_social_presence100 creditsSocial reach and posting cadence: LinkedIn + Facebook follower counts and recent activity.
Ask: “How active is this MSP on social media?”
get_hiring_signals100 creditsActive job postings from Indeed + LinkedIn (title, seniority, recency). Distinguishes “not covered” from “not hiring.”
Ask: “Is this company hiring right now?”
Infrastructure, web & SEO
The technical footprint: email/DNS posture, site performance, and search visibility.
get_infrastructure_profile150 creditsEmail/DNS security posture: SPF/DMARC presence, TLS validity, domain age, and mail/cloud provider flags.
Ask: “How solid is this company's email security?”
get_web_performance100 creditsWebsite performance: Core Web Vitals (field data, lab fallback) and a rolled-up performance band.
Ask: “How fast is this company's website?”
get_seo_profile125 creditsSEO health: audit + web-authority scores and a curated slice of target and ranking keywords.
Ask: “What's this company's SEO health?”
get_serp_visibility150 creditsWhich search queries the company's pages appear for, with position and result type (organic vs. local-pack).
Ask: “What search terms does this MSP show up for?”
Market sizing — counts only
Size a market across the whole database. These return counts only — never company names or lists.
count_companies50 creditsCounts companies matching filters (state, size, service, vendor, maturity signals, recency, and segment — MSP vs VAR…). Returns a total — and an optional breakdown by one dimension, e.g. group by segment for MSPs vs VARs. No names, no lists.
Ask: “How many MSPs vs VARs in Ohio score high on infrastructure security?”
count_social_activity50 creditsCounts how many companies posted on LinkedIn/Facebook — optionally about a topic — in the last N days. No names, no lists.
Ask: “How many MSPs posted about ransomware on LinkedIn in the last 30 days?”
Batch, revenue & changes
Enrich a known list, estimate size, and track what's moved.
get_companies2 credits / match (min 2)Looks up a KNOWN list of up to 25 domains/IDs in one call; returns each match's summary. Batch enrichment of a list you already have — not discovery.
Ask: “Enrich these 20 domains from my list.”
get_revenue_estimate250 creditsA modeled revenue estimate (gross revenue / ARR / MRR / endpoint ranges). A heuristic estimate from benchmarks — not collected data — so its disclaimer is always shown.
Ask: “Roughly how big is this MSP?”
get_recent_changes200 creditsWhat changed recently: rating / review-count / follower deltas plus counts of new reviews, jobs, and posts within a window (default 90 days).
Ask: “What's changed for this company in the last quarter?”
Security, threat & ownership
Exploitable exposure, breach evidence, and who owns whom.
get_security_exposure200 creditsAttack surface (open ports / exposed services) plus CVEs enriched with CISA-KEV (known-exploited) and EPSS exploit probability, and vendor-level technology risk. Evidence is version-unconfirmed — software they run that has actively-exploited CVEs, not proof of a live vulnerability.
Ask: “What's this company's security exposure?”
get_breach_disclosure200 creditsPossible matches of the company to ransomware leak-site listings — evidence only, low-confidence third-party claims surfaced for review, never a confirmed breach. Always shown with its disclaimer.
Ask: “Has this company shown up on any ransomware leak sites?”
get_network_membership200 creditsWhether the company is Independent, part of a PE-backed roll-up, or a franchise — with corporate-family detail (GLEIF ownership, roll-up probability, franchise brand).
Ask: “Is this MSP part of a roll-up or franchise?”
Web, hosting, location & hiring signals
What the site says, where it's hosted, where the company is, and who it's hiring.
get_website_intelligence200 creditsWhat the crawled + AI-classified website says it does and sells: homepage summary, keywords, and whether it advertises pricing, case studies, a team, testimonials, or services, plus per-page summaries.
Ask: “What does this company's website say it offers?”
get_hiring_detail150 creditsFull hiring across every source — ATS boards (Greenhouse / Lever / Ashby / Workable, with the job-description text), plus LinkedIn, Indeed, and Glassdoor. Richer than get_hiring_signals.
Ask: “What roles is this MSP hiring, with the job descriptions?”
get_ip_hosting_profile125 creditsThe primary hosting provider (cloud / CDN / managed host / self-hosted) mapped from the company's resolved public IPs, with ASN detail and geo.
Ask: “Where is this company's infrastructure hosted?”
get_accessibility_audit125 creditsAn axe-core WCAG scan of the homepage: overall pass score, violations by impact, the failed success criteria, and the top failing rules.
Ask: “How accessible is this company's website?”
get_content_velocity100 creditsHow often the company publishes, from its RSS / Atom feeds: posts in the trailing 30 / 90 / 365 days, days since the last post, and a cadence band (active / occasional / dormant / silent). Cadence only — no post content.
Ask: “Is this MSP actively publishing content?”
get_location_profile100 creditsNormalized location: street address, city, state, ZIP, country, headquarters, latitude/longitude, alternate domains, and the Google-Maps address and phone.
Ask: “What's this company's address and phone?”
get_business_density100 creditsHow commercially dense the company's home county is, from US Census County Business Patterns — establishment / employment / payroll counts plus a national percentile and tier.
Ask: “How dense is this company's local market?”
Reference
FAQ
A few things worth knowing before you dig in.
Why do the count tools return numbers, not company lists?
count_companies and count_social_activity are for market sizing — “how big is this segment?” — not lead export. They return totals (and optional breakdowns) but never names or lists. To pull detail on specific companies, look them up individually or with get_companies for a list you already have.
What does “<5” mean in a count?
Very small counts are reported as “<5” rather than an exact number. It protects against pinpointing individual companies from a narrow query — the segment exists, but it's too small to quantify precisely.
Does Instinct return personal contact information?
No. Every tool returns company-level business data only — firmographics, technology, scores, reputation, and market signals. There are no tools that return personal contact details.
How does billing work?
Each tool call debits its credit cost from your balance the moment it runs. Your plan refills that balance to a fixed monthly allotment (use-or-lose). You can cancel anytime — cancellation takes effect at the end of your paid period, so you keep your credits until then.