Access AI-powered MSP market intelligence with company profiles, technographics, technology stacks, reputation, hiring signals, SEO performance, security exposure, market sizing, revenue estimates, and competitive benchmarks.
DeepSeek Harness (DSH) is DeepSeek's agent harness. Each MCP server is one plugin entry in your project config.
This guide connects Instinct MCP to DeepSeek Harness through MewCP, so DeepSeek Harness can use Instinct MCP's 30 tools, like get_company, get_company_profile_full and get_tech_profile, without your Instinct MCP credentials ever reaching the app.
Instinct MCP signs in through managed OAuth. MewCP stores the token encrypted, refreshes it before it expires and injects it only at call time. Your agent never holds it, so it can never leak it.
More apps
Full Google Search Console control for search analytics, sitemaps, sites, and URL inspection, directly from your AI workflows. This MCP server exposes the Google Search Console API for querying search traffic data, managing sitemaps and site properties, and inspecting the indexing status of any URL. It is built for SEO teams monitoring search performance, developers automating sitemap and property management, and analysts pulling search traffic data into reporting pipelines.
Ask your AI to find companies, verify email addresses, and build leads lists with Hunter’s MCP.
Search Apollo's B2B contact and company database, enrich records in bulk, and manage email sequences from your connected Apollo account.
Brevo's marketing and CRM platform — manage contacts, lists and segments, send transactional and campaign email, run SMS and WhatsApp campaigns, and track deals and companies
Other clients
Instinct MCP works the same way in every client below. Pick yours for its own guide.
Free to start.
Scope access to the exact Instinct MCP tools and accounts you allow, work and personal side by side. Anything outside that scope is refused at call time, even when the model asks for it.
Each call is logged: what ran, on which account, and whether it worked. Failures come back structured, so the agent knows whether to retry instead of guessing.
Every app sits behind one secure MCP URL, and there is no server for you to deploy or keep running. Add an integration in the dashboard and it is live on the next call, with no config to change and no new credentials to wire up. Tools load only when a task needs them, so your agent stays fast and focused however many apps you connect.
In your DSH project, or the patch file you mount plugins from.
One plugin instance is one MCP server. Add this to your plugin list.
Tools register as mcp__mewcp__<tool_name>. Sign in when the browser window opens.
Start a new chat and ask: "List the tools you can use from MewCP." You should see your connected apps.
Look up a single company in the Instinct corpus by domain or id, returning its curated profile (name, location, MSP classification).
Retrieve the full enriched firmographic profile for a company (by its company id): description, focus, size, locations, socials, and the composite FCML band.
Retrieve the detected technographic profile for a company (by its company id): the vendor tools and services detected.
Retrieve the FCML score breakdown for a company (by its company id): the composite score/band plus every per-signal sub-score. The premium analytic layer.
Count how many companies in the Instinct corpus match a set of filters. Filters: US state / city / country, size_class, employee_band, fcml_band_min, offered service, detected vendor; a generic signal + signal_band_min over any maturity sub-score (signal one of: website, territory_health, linkedin, facebook, gmaps, indeed, glassdoor, infrastructure_security, tech_stack, compliance, serp_visibility, service_breadth, seo, fcml) — e.g. signal='linkedin', signal_band_min=3 for 'active on LinkedIn'; presence flags has_linkedin/has_facebook/has_google_maps/has_glassdoor; firmographic is_north_american, company_size, founded_after/founded_before; precise recency activity ('jobs' or 'facebook_posts') + active_within_days; and segment ('msp' = MSP (Primary) / bucket 1, 'var' = Managed Services Offered / VAR / bucket 2). Optionally group_by one dimension (state, size_class, country, company_size, is_north_american, fcml_band, segment, or any signal name) — e.g. group_by='segment' for an MSP-vs-VAR breakdown. Returns COUNTS ONLY — never company names or lists — for market sizing / TAM. Small counts are reported as '<5'.
Retrieve a company's NORMALIZED MSP services (by its company id) from the curated services taxonomy — each with its category. Cleaner than free-text technographics.
Retrieve the third-party technology vendors detected for a company (by its company id) — the real technographic signal (RMM/PSA/EDR/backup/…) with category, detection method(s), and confidence.
Retrieve the Territory Health (THS) score for a company (by its company id): the banded local-market strength plus the circle-aggregation explainability (radius, ZIP count, population, urbanicity, anchor).
Retrieve a company's cross-source reputation (by its company id): Google Maps, Glassdoor, and Indeed ratings + review counts, plus recent Glassdoor reviews and Google review metadata.
Retrieve a company's social reach and posting cadence (by its company id): LinkedIn + Facebook follower/employee counts and recent-post activity.
Retrieve a company's active hiring signal (by its company id): current Indeed + LinkedIn job postings (title, seniority, recency) and an active count. Distinguishes 'not covered' from 'not hiring'.
Retrieve a MODELED revenue estimate for a company (by its company id): headline gross-revenue / ARR / MRR / endpoints-under-management ranges. This is a heuristic ESTIMATE from industry benchmarks, NOT collected data — the returned 'disclaimer' must be shown verbatim.
Retrieve a company's email/DNS security posture (by its company id): SPF/DMARC presence, TLS validity, domain age, and mail/cloud provider flags.
Retrieve a company's website performance (by its company id): Core Web Vitals (CrUX field data, lab fallback) and a rolled-up performance band.
Retrieve a company's SEO health (by its company id): rolled-up audit + web-authority scores/bands and a curated slice of target/ranking keywords.
Benchmark a company (by its company id) on one maturity signal against a cohort — returns the company's band + the percentile of the cohort it scores above. signal is one of the sub-scores or 'fcml'. Cohort defaults to the national sellable universe; set cohort_state / cohort_size_class to restrict to the company's own, or cohort_service to a shared service. The cohort auto-widens when too small to benchmark.
Retrieve a company's SERP visibility detail (by its company id): the search queries its pages appear for, with position and result type (organic vs local-pack). The detail behind the serp_visibility score.
Retrieve what changed recently for a company (by its company id): rating / review-count / follower deltas (latest vs previous snapshot) plus counts of new reviews, job posts, and social posts within window_days (default 90).
Look up a KNOWN list of companies in one call — provide up to 25 domains and/or company ids — returning each sellable match's curated summary (name, location, MSP classification). Non-matches are dropped. This is batch enrichment of a known list, NOT discovery. Priced per match returned.
Count how many companies posted on social media recently — 'how many companies posted about <topic> on LinkedIn/Facebook in the last N days'. Params: network ('linkedin' | 'facebook' | 'any'), within_days, and an optional topic (full-text search over post text; omit for 'posted at all'). Returns COUNTS ONLY — the number of distinct companies (and total matching posts) — never company names or lists. Small counts are reported as '<5'.
Retrieve a company's security exposure (by its company id): passive attack surface (open ports / exposed services from Shodan InternetDB), CPE-inferred 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.
Retrieve a company's ransomware leak-site exposure (by its company id) — EVIDENCE ONLY. Returns possible matches of the company to a ransomware group's leak-site listing (matched by registrable domain or exact distinctive name), each a low-confidence third-party claim surfaced for review, NEVER a confirmed breach. The 'disclaimer' field MUST be rendered verbatim.
Retrieve a company's network-membership label (by its company id): Independent vs PE-roll-up vs Franchise, with corporate-family detail (GLEIF ownership graph, fused roll-up probability, corporate parent/LEI) and franchise brand. 'network_type' is a conservative derived label. Ownership evidence is GLEIF (CC0); not a scored dimension.
Retrieve the business density of a company's home county (by its company id) from US Census County Business Patterns: establishment / employment / payroll counts plus national percentile and density tier. Territory context, not a scored dimension.
Return a company's hosting / IP-network profile (by its company id), derived from the Instinct corpus' resolved public IP addresses: the primary hosting provider (cloud, CDN, managed host, or self-hosted) mapped from each IP's ASN, cloud/CDN/self-hosted flags, per-IP ASN detail, and geo. Returns available=false when the company has no resolved public IPs on file.
Return a company's web accessibility audit (by its company id) from an axe-core WCAG scan of its homepage in the Instinct corpus: the overall pass score, violation counts by impact, the failed WCAG success criteria, and the top failing rules. Returns available=false when no accessibility scan is on file.
Return a company's content-publishing cadence (by its company id) from its RSS / Atom feeds in the Instinct corpus: posts in the trailing 30 / 90 / 365 days, days since the last post, median interval, feed count, and a coarse cadence band (active / occasional / dormant / silent). Cadence metadata only, never post content. Returns available=false when no feed was discovered.
Return a company's normalized location (by its company id) from the Instinct corpus: structured street address, city, state, ZIP, country, headquarters, territory-anchor latitude/longitude, alternate domains, and the Google-Maps address and phone. Business contact detail only. Returns available=false when no location detail is on file.
Return a company's website intelligence (by its company id) from the Instinct corpus' crawled + AI-classified pages: the homepage summary, keywords, whether the site advertises pricing / case studies / team / testimonials / services, and a bounded set of page summaries. Long free-text is returned as tagged untrusted content. Returns available=false when no pages are on file.
Return a company's full hiring detail (by its company id) from the Instinct corpus, unified across ATS boards (Greenhouse / Lever / Ashby / Workable, with full job-description text), LinkedIn, Indeed, and Glassdoor. Each posting includes source, title, department, location, seniority, and posted date; ATS postings also include the job-description text (tagged untrusted content). Richer than get_hiring_signals. Returns is_hiring=false when no open postings are found.
