Workforce Competitive Intelligence MCP Server

Workforce competitive intelligence at query speed — this MCP server gives your AI assistant live access to hiring signals, patent inventor movement, researcher attrition, technical capability maps, and executive flight risk for any company. Connect it to Claude, Cursor, or any MCP-compatible client and ask questions that used to require a full analyst team.

The server orchestrates 7 independent data sources — job postings, USPTO patents, EPO patents, ORCID researcher profiles, company research, GitHub activity, and SEC Form 4 insider trading filings — running them in parallel and applying four scoring models to produce structured workforce intelligence. Every tool call returns a scored, signal-annotated JSON response with INVEST/MONITOR/CAUTION/AVOID verdicts. No subscription. No spreadsheet. Pay $0.045 per query.

What data can you extract?

Why use Workforce Competitive Intelligence MCP Server?

Building a workforce intelligence picture manually means pulling job listings from five sources, cross-referencing USPTO and EPO patent filings, checking ORCID profiles, reviewing SEC EDGAR Form 4 filings, and combing GitHub for engineering signals. For a single company that is a full day's work. For a competitor comparison it is two days. For a portfolio of ten companies it simply does not happen.

This MCP server automates the entire process. Ask your AI assistant a single question — "What is the human capital risk for Palantir?" — and receive a scored, signal-annotated dossier in under two minutes, drawing on all 7 sources at once.

• Scheduling — run recurring workforce monitors daily or weekly so your AI receives fresh hiring signals without any manual trigger
• API access — trigger analyses from Python, JavaScript, or any HTTP client alongside your existing research pipelines
• Parallel execution — all 7 actor calls run concurrently so a full dossier completes in the time of a single sequential lookup
• Spending controls — set a per-run maximum charge so cost never exceeds what you authorize
• Integrations — connect results to Zapier, Make, webhooks, or your CRM for automated talent risk alerting

Features

• Talent Velocity Score (0-100) — composite of job posting volume (max 40 pts), hiring function diversity via HHI analysis (max 25 pts), GitHub activity from the last 180 days (max 20 pts), and company growth signals (max 15 pts)
• Brain Drain Index (0-100) — patent inventor continuity analysis across USPTO and EPO filings, ORCID multi-affiliation mobility rate, SEC insider sell ratio as a leadership flight proxy, and low-hiring combined signals
• Competitive Capability Map — maps 10 technology domains (AI/ML, Cloud, Blockchain, Cybersecurity, Mobile, Data/Analytics, Frontend, Backend, DevOps) across job titles, patent titles, and GitHub topics using keyword frequency scoring
• Executive Flight Risk (0-100) — SEC Form 4 sell/buy ratio (max 40 pts), serial seller concentration for named executives (max 20 pts), C-suite replacement hiring (CEO/CTO/CFO/COO/VP openings, max 20 pts), company distress text signals (max 20 pts)
• Composite Workforce Dossier — weighted composite of Talent Velocity (25%), inverted Brain Drain (25%), Competitive Capability (30%), and inverted Executive Flight Risk (20%) with INVEST/MONITOR/CAUTION/AVOID override logic
• AVOID override — if Executive Flight Risk is CRITICAL or Brain Drain is HEMORRHAGING the verdict is forced to AVOID regardless of composite score
• 7 parallel data sources — job postings, USPTO patents, EPO patents, ORCID researchers, company research, GitHub repos, and SEC insider transactions fetched simultaneously
• Signal narration — every score includes human-readable signal strings (e.g., "9 senior/leadership roles — building new org layers") for direct use in AI-generated reports
• Two-company benchmarking — benchmark_talent_strategy runs both companies in parallel and returns side-by-side Talent Velocity and Competitive Capability scores with a declared advantage winner
• Human capital risk composite — assess_human_capital_risk weights Brain Drain at 55% and Executive Flight Risk at 45% for a single organizational health number
• Growth signal classification — five-tier labels: CONTRACTING, STABLE, GROWING, SURGING, HYPERGROWTH applied to Talent Velocity output
• Drain level classification — five-tier labels: RETAINING, STABLE, AT_RISK, DRAINING, HEMORRHAGING applied to Brain Drain output
• Capability level classification — five-tier labels: NASCENT, DEVELOPING, COMPETITIVE, LEADING, DOMINANT applied to Competitive Capability output
• Pay-per-event pricing — $0.045 per tool call with no standing subscription

Use cases for workforce competitive intelligence

Competitor strategic pivot detection

Corporate strategy teams and competitive intelligence analysts need to know when a competitor is shifting direction months before the press release. An AI assistant connected to this MCP can answer "Is Palantir pivoting toward defense AI hiring?" with live job posting volume, seniority mix, and function distribution data. Sudden engineering concentration in AI/ML roles, combined with US and EU patent filings in the same domain, provides a 3-6 month early signal ahead of public announcements.

M&A human capital due diligence

Investment bankers, PE analysts, and corp dev teams running acquisition due diligence need to quantify human capital risk before close. The generate_workforce_dossier tool produces a scored dossier with INVEST/MONITOR/CAUTION/AVOID verdicts in under two minutes. Pair this with patent inventor movement data to assess whether key IP creators are still active or have departed for competitors, directly affecting the target's intangible asset valuation.

Executive retention early warning

Investor relations teams, board members, and HR leaders at public companies track SEC Form 4 insider trading as an executive stability signal. The monitor_executive_transitions tool identifies executives with high sell ratios, large individual transactions above $1M, and unusual concentration patterns (serial sellers) and flags C-suite replacement roles open simultaneously — a multi-signal indicator of imminent leadership instability.

Talent strategy benchmarking

Heads of talent acquisition and CHROs comparing their organization against a direct competitor need more than LinkedIn headcount numbers. The benchmark_talent_strategy tool runs both companies through parallel job posting, patent, and GitHub analysis and returns side-by-side Talent Velocity Scores and Competitive Capability Maps, identifying which company has the stronger hiring momentum and deeper technical domain presence.

Academic-to-industry talent monitoring

Research-intensive industries — pharma, defense, semiconductor, biotech — need to know when key academic researchers affiliate with commercial competitors. The detect_researcher_attrition tool queries ORCID profiles for multi-affiliation researchers (scientists with two or more institutional ties), cross-references active patent filings, and calculates a mobility rate that signals potential technology transfer risks.

Portfolio talent risk monitoring

Venture capital and private equity firms with 10+ portfolio companies cannot manually track workforce health for every holding. Schedule assess_human_capital_risk runs weekly for each company and route results via webhook to a Slack channel or CRM. Flag any holding that crosses an 60+ human capital risk threshold for proactive founder conversations before talent issues compound.

How to connect this MCP server

Step 1: Get your Apify API token

Sign up at apify.com and copy your token from Settings > Integrations. The Apify free plan includes $5 of monthly credits — enough for roughly 111 tool calls.

Step 2: Add the MCP server to your client

The server endpoint is:

https://workforce-competitive-intelligence-mcp.apify.actor/mcp

Use your Apify token for authentication.

Step 3: Configure your MCP client

Claude Desktop — add to claude_desktop_config.json:

{
  "mcpServers": {
    "workforce-competitive": {
      "url": "https://workforce-competitive-intelligence-mcp.apify.actor/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_APIFY_TOKEN"
      }
    }
  }
}

Cursor / Windsurf / Cline — add the server URL and Authorization header in your MCP settings panel. The server responds to standard MCP protocol over HTTP POST.

Step 4: Start querying

Ask your AI assistant natural language questions:

• "Generate a full workforce dossier for Stripe"
• "Compare hiring velocity between Snowflake and Databricks"
• "What is the executive flight risk for Intel?"
• "Track inventor movement at NVIDIA"

MCP tools

Tool tips

• Use generate_workforce_dossier for unknown targets — it pulls all 7 sources and all 4 models, giving you the widest picture before narrowing with individual tools
• Use benchmark_talent_strategy before a competitive pitch — side-by-side scores give your team a defensible, quantified talking point
• Use assess_human_capital_risk for portfolio monitoring — it is the cheapest path to a single risk number suitable for threshold-based alerting
• Use track_inventor_movement for IP due diligence — the top inventors list shows exactly who has been creating IP and whether their output is declining
• Chain tools for deeper analysis — run monitor_executive_transitions first and if flight risk is HIGH, follow with detect_researcher_attrition to confirm whether the leadership instability is correlating with talent loss

Output example

The following is a representative response from generate_workforce_dossier for a mid-size technology company:

{
  "entity": "Meridian Analytics Inc",
  "compositeScore": 68,
  "verdict": "MONITOR",
  "talentVelocity": {
    "score": 74,
    "totalOpenings": 38,
    "seniorRoles": 11,
    "techRoles": 22,
    "growthSignal": "SURGING",
    "signals": [
      "38 open positions — aggressive hiring campaign",
      "11 senior/leadership roles — building new org layers",
      "22 technical roles — major engineering investment",
      "Hiring across 5 functions — broad organizational growth"
    ]
  },
  "brainDrain": {
    "score": 37,
    "inventorCount": 14,
    "researcherCount": 11,
    "patentActivity": 19,
    "drainLevel": "AT_RISK",
    "signals": [
      "4 researchers with multiple affiliations — talent mobility risk"
    ]
  },
  "competitiveCapability": {
    "score": 71,
    "techDomains": ["AI/ML", "Cloud", "Data/Analytics", "DevOps", "Backend"],
    "patentStrength": 19,
    "talentDepth": 24,
    "capabilityLevel": "LEADING",
    "signals": [
      "Active across 5 tech domains: AI/ML, Cloud, Data/Analytics, DevOps, Backend",
      "19 patents — strong IP portfolio",
      "Patents in both US and EU — global IP strategy",
      "24 public repos — significant open-source investment"
    ]
  },
  "executiveFlight": {
    "score": 29,
    "insiderSells": 4,
    "insiderBuys": 7,
    "sellRatio": 0.36,
    "riskLevel": "MODERATE",
    "signals": []
  },
  "allSignals": [
    "38 open positions — aggressive hiring campaign",
    "11 senior/leadership roles — building new org layers",
    "22 technical roles — major engineering investment",
    "Active across 5 tech domains: AI/ML, Cloud, Data/Analytics, DevOps, Backend",
    "19 patents — strong IP portfolio",
    "4 researchers with multiple affiliations — talent mobility risk"
  ],
  "keyRisks": [],
  "keyStrengths": [
    "Strong talent velocity — aggressive hiring signals",
    "Strong competitive position — deep IP and tech capabilities",
    "Executive stability — insiders holding/buying"
  ]
}

Output fields

How much does it cost to run workforce intelligence queries?

This MCP server uses pay-per-event pricing — each tool call costs $0.045. Platform compute costs are included. You pay nothing while the server is idle.

You can set a maximum spending limit per run to control costs. The server stops charging when your budget is reached and returns a clear limit-reached message.

Compare this to dedicated talent intelligence platforms charging $15,000-80,000/year for comparable workforce signal data. Most users of this MCP spend $2-20/month with no subscription commitment.

Using the API directly

Python

import requests
import json

token = "YOUR_APIFY_TOKEN"
url = "https://workforce-competitive-intelligence-mcp.apify.actor/mcp"
headers = {
    "Content-Type": "application/json",
    "Authorization": f"Bearer {token}"
}

payload = {
    "jsonrpc": "2.0",
    "method": "tools/call",
    "params": {
        "name": "generate_workforce_dossier",
        "arguments": {"company": "Snowflake"}
    },
    "id": 1
}

response = requests.post(url, headers=headers, json=payload)
result = response.json()
content = json.loads(result["result"]["content"][0]["text"])

print(f"Company: {content['entity']}")
print(f"Verdict: {content['verdict']} (composite score: {content['compositeScore']})")
print(f"Growth signal: {content['talentVelocity']['growthSignal']}")
print(f"Drain level: {content['brainDrain']['drainLevel']}")
print(f"Capability: {content['competitiveCapability']['capabilityLevel']}")
print(f"Executive risk: {content['executiveFlight']['riskLevel']}")
for signal in content["allSignals"]:
    print(f"  - {signal}")

JavaScript

const token = "YOUR_APIFY_TOKEN";
const url = "https://workforce-competitive-intelligence-mcp.apify.actor/mcp";

const response = await fetch(url, {
  method: "POST",
  headers: {
    "Content-Type": "application/json",
    "Authorization": `Bearer ${token}`
  },
  body: JSON.stringify({
    jsonrpc: "2.0",
    method: "tools/call",
    params: {
      name: "benchmark_talent_strategy",
      arguments: { company_a: "Databricks", company_b: "Snowflake" }
    },
    id: 1
  })
});

const result = await response.json();
const content = JSON.parse(result.result.content[0].text);

const { comparison, talentAdvantage, capabilityAdvantage } = content;
console.log(`Talent velocity advantage: ${talentAdvantage}`);
console.log(`Capability advantage: ${capabilityAdvantage}`);

for (const [company, data] of Object.entries(comparison)) {
  const d = data as any;
  console.log(`${company}: velocity ${d.talentVelocity.score} (${d.talentVelocity.growthSignal}), capability ${d.competitiveCapability.score} (${d.competitiveCapability.capabilityLevel})`);
}

cURL

# Run a full workforce dossier
curl -X POST "https://workforce-competitive-intelligence-mcp.apify.actor/mcp" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_APIFY_TOKEN" \
  -d '{
    "jsonrpc": "2.0",
    "method": "tools/call",
    "params": {
      "name": "generate_workforce_dossier",
      "arguments": {"company": "Palantir Technologies"}
    },
    "id": 1
  }'

# Run a two-company benchmark
curl -X POST "https://workforce-competitive-intelligence-mcp.apify.actor/mcp" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_APIFY_TOKEN" \
  -d '{
    "jsonrpc": "2.0",
    "method": "tools/call",
    "params": {
      "name": "benchmark_talent_strategy",
      "arguments": {"company_a": "OpenAI", "company_b": "Anthropic"}
    },
    "id": 2
  }'

How Workforce Competitive Intelligence MCP works

Phase 1: Parallel data collection across 7 sources

When a tool is called, the server dispatches up to 7 Apify actor calls simultaneously using Promise.allSettled. Each actor runs independently at 256 MB memory with a 120-second timeout. If any individual actor fails, the remaining results are still returned and that source's data is treated as an empty array — so a USPTO outage does not block a hiring analysis. The 7 actors are: job-market-intelligence, patent-search (USPTO), epo-patent-search (EPO), orcid-researcher-search, company-deep-research, github-repo-search, and sec-insider-trading.

Phase 2: Scoring model application

The four scoring models run against the collected data arrays:

Talent Velocity uses a point-ceiling model. Job volume contributes up to 40 points (volume * 2 + senior roles * 3, capped at 40). Function diversity uses a Herfindahl-Hirschman Index (HHI) — lower concentration produces higher diversity scores up to 25 points. GitHub activity from the last 180 days contributes up to 20 points. Company growth signals (funding, expansion, headcount keywords) contribute up to 15 points.

Brain Drain scores patent continuity by comparing recent-year filing counts against older filings — a 50%+ decline triggers a 25-point penalty signal. ORCID multi-affiliation rate (researchers with 2+ institutional ties) contributes up to 25 points. SEC sell ratio contributes up to 25 points. A combined low-hiring-plus-high-selling signal adds up to 15 points.

Competitive Capability extracts technology domain presence using a 10-domain keyword dictionary applied across job title text, patent title text, and GitHub topic/description text. Domain count scores up to 30 points. Total patent count scores up to 30 points (2 points per patent). GitHub depth (repo count + stars + contributors) scores up to 25 points. Cross-domain hiring breadth contributes up to 15 points.

Executive Flight Risk analyzes Form 4 transaction types. The sell/buy ratio drives the primary score (up to 40 points), amplified by large individual sales above $1M (5 points each). Serial sellers — executives with 3+ sell transactions — are named individually in signals and contribute up to 20 points via a concentration score. C-suite replacement roles (CEO, CTO, CFO, COO, SVP) in job postings score 5 points each, capped at 20. Company distress keywords (layoff, restructuring, lawsuit, SEC probe) contribute up to 20 points.

Phase 3: Composite assembly and verdict

The workforce dossier composite weights are: Talent Velocity 25%, inverted Brain Drain 25% (100 - drain score), Competitive Capability 30%, inverted Executive Flight Risk 20%. Two hard override rules apply regardless of the composite score: if Executive Flight Risk is CRITICAL (score >= 80) or Brain Drain is HEMORRHAGING (score >= 80), the verdict is forced to AVOID.

Phase 4: Signal narration and structured output

All scoring functions generate human-readable signal strings when thresholds are crossed — for example, "47 technical roles — major engineering investment" or "3 large sales (>$1M) — significant executive divestment." The dossier aggregates all signals from all four models into a single allSignals array and separates keyRisks and keyStrengths based on sub-score threshold breaches.

Tips for best results

1. Start with generate_workforce_dossier for any new target. It uses all 7 sources and returns the composite picture, key risks, and key strengths in a single call. Use individual tools afterward only to go deeper on a specific dimension.

2. For public companies, executive flight risk is most actionable. SEC Form 4 data is precise and timestamped. A sell ratio above 0.7 with 3+ transactions is a quantifiable signal you can bring into an investment committee.

3. Treat Brain Drain Index as a lagging indicator. Patent filings take months to process and ORCID profiles are updated sporadically. A score of AT_RISK or higher confirms a pattern, but the talent may have departed 6-12 months earlier. Pair with Talent Velocity (a leading indicator) for the complete timeline.

4. Use benchmark_talent_strategy before a board presentation. The tool returns which company has the talent velocity advantage and which has the capability advantage — two ready-made data points that do not require further explanation.

5. Set a spending limit for portfolio monitoring. If you are running weekly checks across 20 companies, set a per-session budget cap in your Apify account to prevent unexpected charges if a batch grows.

6. Combine inventor tracking with capability mapping. Run track_inventor_movement and map_technical_capabilities on the same entity to correlate which tech domains are losing inventors. A domain where patents are declining and inventors are departing is a genuine IP risk, not just a filing timing artifact.

7. For private companies, expect lower SEC and patent coverage. Private companies do not file Form 4 reports and may file fewer patents. In those cases, the Talent Velocity score and GitHub depth carry more analytical weight than the Brain Drain or Executive Flight scores.

Combine with other Apify actors

Limitations

• Private company SEC coverage is nil. SEC Form 4 filings are only required from officers and directors of public companies. Executive Flight Risk scores for private companies rely entirely on the hiring and company research signals, which are weaker proxies.
• Patent data lags by weeks to months. USPTO and EPO patent publication timelines mean recent inventor departures may not yet be reflected in the patent record. The Brain Drain Index is more reliable as a confirmatory signal than an early warning system.
• Job posting data reflects advertised roles only. Companies that hire primarily via referral, internal promotion, or recruiter-only pipelines will appear to have lower Talent Velocity scores than their actual hiring pace.
• ORCID coverage skews academic. Researchers who do not maintain public ORCID profiles — which includes most commercial R&D employees — are not captured. The researcher attrition signal is most reliable for companies with strong academic-industry research programs.
• GitHub analysis reflects open-source output only. Companies with significant proprietary engineering activity and minimal public GitHub presence (common in defense, finance, and enterprise software) will score lower on technical depth than their actual capability warrants.
• The 120-second actor timeout may truncate large results. For very active companies with thousands of job postings or hundreds of patents, results are limited by what each underlying actor returns within the timeout window.
• Company name disambiguation is not performed. Passing "Apple" retrieves data for Apple Inc but may mix in results for other entities named Apple. Use full legal names or include country/ticker qualifiers for precision.
• Composite verdicts are quantitative signals, not investment advice. INVEST/MONITOR/CAUTION/AVOID verdicts reflect aggregated public data signals and should be one input among many in any investment or strategic decision.

Integrations

• Apify API — call any tool programmatically from Python, JavaScript, or any HTTP client for automated workforce monitoring pipelines
• Zapier — trigger workforce dossier runs on a schedule and route CAUTION/AVOID verdicts to Slack or email for portfolio risk alerts
• Make — build automated competitive intelligence workflows that run weekly talent velocity checks and write results to Google Sheets
• Google Sheets — append dossier verdicts and composite scores to a competitive tracking spreadsheet updated automatically after each run
• Webhooks — fire a webhook when any company's Brain Drain or Executive Flight Risk score exceeds a configured threshold
• LangChain / LlamaIndex — embed workforce intelligence as a tool in LangChain or LlamaIndex agent workflows for autonomous competitive research

Troubleshooting

Returning empty signals despite an active company. Company name disambiguation is performed by the underlying actors. If you pass "Microsoft" you will get strong results, but niche company names may return sparse data. Try including the full legal name, country qualifier, or stock ticker. For track_inventor_movement, try the company's primary product name or technology area rather than the corporate name.

Brain Drain score unexpectedly high for a healthy company. Patent filing gaps are common due to processing delays at USPTO and EPO, which can trigger the patent decline signal even when inventors are active. Cross-check with detect_researcher_attrition — if researcher mobility is low, the patent signal is likely a filing timing artifact rather than genuine talent loss.

generate_workforce_dossier taking longer than expected. The full dossier runs 7 actor calls in parallel, each with a 120-second timeout. Total wall time is typically 60-120 seconds. If a run exceeds 3 minutes, check the Apify actor logs for individual actor timeouts. Reducing to a single-source tool (e.g., analyze_hiring_signals) will return faster for exploratory queries.

Executive Flight Risk showing CRITICAL for a stable company. A single quarter with heavy insider selling (common after lock-up expirations or planned 10b5-1 trading programs) can elevate the sell ratio temporarily. Review the insiderTransactions array in the raw response to distinguish programmatic selling plans from unplanned divestment patterns.

Token authentication failing. The server requires the Authorization header with Bearer YOUR_APIFY_TOKEN. Some MCP clients pass tokens differently. If using Claude Desktop, confirm the header format in your claude_desktop_config.json matches the example above. Your token is available at console.apify.com/account/integrations.

Responsible use

• This server queries publicly available government databases (USPTO, EPO, SEC EDGAR), open researcher profiles (ORCID), and public software repositories (GitHub).
• Patent inventor data and SEC filing data name individuals in their professional capacity as inventors and corporate officers — this is public disclosure mandated by law.
• ORCID profiles are created and maintained by researchers who explicitly opted into public visibility.
• Do not use workforce intelligence outputs to target individual employees for unsolicited contact.
• Comply with applicable data protection regulations when incorporating this data into HR or recruiting decisions.
• For guidance on data use legality, see Apify's guide on web scraping legality.

FAQ

How does workforce competitive intelligence differ from LinkedIn Sales Navigator or Talent Insights? LinkedIn provides headcount snapshots and basic role distribution. This MCP adds dimensions that LinkedIn cannot: patent inventor continuity (USPTO/EPO), researcher affiliation mobility (ORCID), executive stock transaction patterns (SEC Form 4), and GitHub engineering depth. These alternative signals often lead LinkedIn data by months because patents and ORCID updates capture movement that has not yet been reflected in profile updates.

How accurate is the Talent Velocity Score for smaller companies? Smaller companies (under 200 employees) typically have fewer open job listings and GitHub repos, so scores trend lower regardless of actual hiring pace. The signal is more reliable as a comparative tool than as an absolute measure — use benchmark_talent_strategy to compare a target against a peer of similar size rather than reading the absolute score in isolation.

Can workforce competitive intelligence track individual employees? The server tracks patent inventors (named on public USPTO and EPO filings), ORCID-registered researchers (who opted into public profiles), and corporate officers named in SEC Form 4 filings. It does not track general employees and does not access private employment records or LinkedIn profiles.

How current is the job posting data? Job postings are fetched live at query time from the job market intelligence actor, reflecting currently active listings. Patent filings lag by weeks to months depending on USPTO and EPO processing timelines. SEC Form 4 data is typically available within 2 business days of a transaction. ORCID data reflects the most recent researcher profile update, which varies by individual.

Is it legal to analyze competitor hiring data and SEC filings this way? Yes. Job postings are public advertisements. SEC Form 4 filings are legally mandated public disclosures. Patent filings are public records by design to enable knowledge dissemination. ORCID profiles are voluntarily made public by researchers. For a detailed analysis of data use legality, see Apify's guide.

What does the AVOID verdict actually mean? AVOID is triggered either by a composite score below 30 or by a hard override: Executive Flight Risk classified as CRITICAL (score >= 80) or Brain Drain classified as HEMORRHAGING (score >= 80). It indicates significant, multi-signal human capital instability. In an M&A context, it flags targets that require deep retention planning before close. In a competitive context, it identifies competitors potentially entering a period of strategic weakness.

How does executive flight risk detection work in practice? The model analyzes SEC Form 4 transaction type classifications. It counts sell vs buy transactions, identifies executives with 3+ individual sell events (serial sellers), flags transactions above $1M, and checks for simultaneous C-suite job openings. A sell ratio above 0.8 with 3+ transactions triggers a "executives liquidating positions" signal. This pattern has historically preceded executive departures, particularly when paired with distress keywords in company research.

Can I schedule this to monitor a list of companies automatically? Yes. You can schedule recurring Apify actor runs via the Apify platform scheduler. Use the HTTP API to call specific tools on a list of company names and route results via webhook to Slack, email, or a CRM. The assess_human_capital_risk tool is the most cost-efficient option for recurring monitoring ($0.045 per company per run).

Does this work with Claude Desktop's MCP integration? Yes. Add the server URL and Authorization header to your claude_desktop_config.json as shown in the connection guide above. Once connected, you can ask Claude natural language questions about any company and it will call the appropriate tool, interpret the scores, and explain the signals in plain language.

What happens if one of the 7 underlying data sources is unavailable? The server uses Promise.allSettled for parallel actor execution, which means individual actor failures do not block the overall response. If, for example, the EPO actor times out, the response still includes all other source data and the scoring models treat the missing source as an empty array. The returned signals will note limited coverage for that dimension.

How does the server handle company name ambiguity? It does not perform entity disambiguation automatically. The company name you pass is used directly as a search query to each underlying actor. For common names (e.g., "Oracle", "Amazon") this works well. For ambiguous names, append the country, industry, or stock ticker to the query — for example "Palantir Technologies NYSE" rather than just "Palantir".

What MCP clients work with this server? Any client that supports the Model Context Protocol over HTTP POST works: Claude Desktop, Cursor, Windsurf, Cline, and custom LangChain or LlamaIndex agents. The server uses the StreamableHTTPServerTransport from the official MCP SDK and responds at the /mcp endpoint.

Help us improve

If you encounter issues, you can help us debug faster by enabling run sharing in your Apify account:

1. Go to Account Settings > Privacy
2. Enable Share runs with public Actor creators

This lets us see your run details when something goes wrong, so we can fix issues faster. Your data is only visible to the actor developer, not publicly.

Support

Found a bug or have a feature request? Open an issue in the Issues tab on this actor's page. For custom enterprise integrations or additional data source coverage, reach out through the Apify platform.