@mgsoftwarebv/mg-dashboard-mcp 2.4.3 → 2.5.1

package/dist/index.js

@@ -413,6 +413,9 @@ var VALID_REVIEW_STATUSES = /* @__PURE__ */ new Set([
   "agent_flagged",
   "human_approved"
 ]);
+function normalizeCompanyName(name) {
+  return name.toLowerCase().replace(/\b(b\.?v\.?|n\.?v\.?|v\.?o\.?f\.?|c\.?v\.?|holding|groep|group|nederland|netherlands)\b/gi, "").replace(/[^a-z0-9\s]/g, "").replace(/\s+/g, " ").trim();
+}
 var AGENT_TOOLS = [
   {
     name: "agent-report-coverage",
@@ -632,6 +635,103 @@ var AGENT_TOOLS = [
       },
       required: ["repo_slug", "results"]
     }
+  },
+  // -- Lead Generation tools --------------------------------------------------
+  {
+    name: "agent-check-lead-exists",
+    description: "Check if a lead already exists by website URL or company name. Call this BEFORE visiting a website to avoid wasting time on duplicates.",
+    inputSchema: {
+      type: "object",
+      properties: {
+        website_url: {
+          type: "string",
+          description: 'Website URL to check (e.g. "https://example.nl")'
+        },
+        company_name: {
+          type: "string",
+          description: "Company name to fuzzy-match against existing leads"
+        }
+      },
+      required: []
+    }
+  },
+  {
+    name: "agent-save-lead",
+    description: "Save a discovered lead (company) to the database. Handles dedup on website_url + fuzzy match on company name. Returns the lead ID and whether it already existed.",
+    inputSchema: {
+      type: "object",
+      properties: {
+        company_name: { type: "string", description: "Company name" },
+        website_url: { type: "string", description: "Company website URL" },
+        industry: { type: "string", description: "Industry sector" },
+        region: { type: "string", description: 'Geographic region (e.g. "amsterdam", "rotterdam")' },
+        description: { type: "string", description: "AI-generated summary of what the company does (max 2000 chars)" },
+        potential_fit: { type: "string", description: "Why MG Software could help this company (max 2000 chars)" },
+        fit_score: { type: "number", description: "Fit score 1-10 for MG Software partnership" },
+        estimated_company_size: { type: "string", description: 'Estimated employee count range (e.g. "1-10", "10-50", "50-200")' },
+        kvk_number: { type: "string", description: "KvK (Chamber of Commerce) number if found" },
+        contact_name: { type: "string", description: "Primary contact person name" },
+        contact_role: { type: "string", description: "Contact person role/title" },
+        contact_email: { type: "string", description: "Contact person email" },
+        contact_phone: { type: "string", description: "Contact person phone" },
+        contact_linkedin: { type: "string", description: "Contact person LinkedIn URL" },
+        general_email: { type: "string", description: "General company email (info@...)" },
+        general_phone: { type: "string", description: "General company phone number" },
+        source_url: { type: "string", description: "URL where this lead was found (Google result, directory, etc.)" },
+        target_id: { type: "string", description: "UUID of the lead_generation_target this lead was found for" }
+      },
+      required: ["company_name", "website_url"]
+    }
+  },
+  {
+    name: "agent-save-email-draft",
+    description: "Save a cold email draft for a lead. The email will be shown in the backoffice for manual review and copy. Do NOT use em-dashes or en-dashes in the email.",
+    inputSchema: {
+      type: "object",
+      properties: {
+        lead_id: { type: "string", description: "UUID of the lead this email is for" },
+        subject: { type: "string", description: "Email subject line (Dutch, max 200 chars)" },
+        body: { type: "string", description: "Email body text (Dutch, max 5000 chars, NO em-dashes or en-dashes)" },
+        tone: { type: "string", description: "Tone of the email (default: professional)" }
+      },
+      required: ["lead_id", "subject", "body"]
+    }
+  },
+  {
+    name: "agent-complete-target",
+    description: "Mark a lead_generation_target as completed with the number of leads found.",
+    inputSchema: {
+      type: "object",
+      properties: {
+        target_id: { type: "string", description: "UUID of the lead_generation_target" },
+        results_count: { type: "number", description: "Number of new leads saved for this target" }
+      },
+      required: ["target_id", "results_count"]
+    }
+  },
+  {
+    name: "web-search",
+    description: "Search the web using DuckDuckGo. Returns a list of results with title, URL, and snippet. Use this to find companies, websites, directories, etc.",
+    inputSchema: {
+      type: "object",
+      properties: {
+        query: { type: "string", description: 'Search query (e.g. "logistiek bedrijf MKB nederland")' },
+        max_results: { type: "number", description: "Max results to return (default: 15, max: 30)" }
+      },
+      required: ["query"]
+    }
+  },
+  {
+    name: "web-fetch",
+    description: "Fetch a web page and return its text content (HTML tags stripped). Use this to read company websites, extract contact info, about pages, etc. Returns at most 15000 characters of cleaned text.",
+    inputSchema: {
+      type: "object",
+      properties: {
+        url: { type: "string", description: 'Full URL to fetch (e.g. "https://example.nl/contact")' },
+        extract_links: { type: "boolean", description: "Also extract all links from the page (default: false)" }
+      },
+      required: ["url"]
+    }
   }
 ];
 var AGENT_TOOL_NAMES = new Set(AGENT_TOOLS.map((t) => t.name));
@@ -641,7 +741,13 @@ var AGENT_TOOL_MODULE_MAP = {
   "agent-save-documentation": "agent_reporting",
   "agent-list-findings": "agent_reporting",
   "agent-get-documentation": "agent_reporting",
-  "agent-validate-suggestions": "agent_reporting"
+  "agent-validate-suggestions": "agent_reporting",
+  "agent-check-lead-exists": "agent_reporting",
+  "agent-save-lead": "agent_reporting",
+  "agent-save-email-draft": "agent_reporting",
+  "agent-complete-target": "agent_reporting",
+  "web-search": "agent_reporting",
+  "web-fetch": "agent_reporting"
 };
 function clamp(val, min, max) {
   return Math.max(min, Math.min(max, val));
@@ -649,6 +755,80 @@ function clamp(val, min, max) {
 function sanitizeString(val, maxLen) {
   return String(val ?? "").slice(0, maxLen);
 }
+var WEB_USER_AGENT = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36";
+async function webSearch(query, maxResults) {
+  const url = `https://html.duckduckgo.com/html/?q=${encodeURIComponent(query)}`;
+  const response = await fetch(url, {
+    headers: {
+      "User-Agent": WEB_USER_AGENT,
+      "Accept": "text/html",
+      "Accept-Language": "nl,en;q=0.9"
+    },
+    redirect: "follow"
+  });
+  if (!response.ok) throw new Error(`Search failed: HTTP ${response.status}`);
+  const html = await response.text();
+  const results = [];
+  const resultBlocks = html.split(/class="result\s/);
+  for (let i = 1; i < resultBlocks.length && results.length < maxResults; i++) {
+    const block = resultBlocks[i];
+    const titleMatch = block.match(/class="result__a"[^>]*>([^<]+)</);
+    const hrefMatch = block.match(/class="result__a"\s+href="([^"]+)"/);
+    const snippetMatch = block.match(/class="result__snippet"[^>]*>([\s\S]*?)<\/a>/);
+    if (!hrefMatch) continue;
+    let href = hrefMatch[1];
+    if (href.includes("uddg=")) {
+      const uddg = new URL(href, "https://duckduckgo.com").searchParams.get("uddg");
+      if (uddg) href = uddg;
+    }
+    if (!href.startsWith("http")) continue;
+    const title = titleMatch ? titleMatch[1].replace(/&amp;/g, "&").replace(/&lt;/g, "<").replace(/&gt;/g, ">").replace(/&#x27;/g, "'").replace(/&quot;/g, '"').trim() : href;
+    const snippet = snippetMatch ? snippetMatch[1].replace(/<[^>]+>/g, "").replace(/&amp;/g, "&").replace(/&lt;/g, "<").replace(/&gt;/g, ">").replace(/&#x27;/g, "'").replace(/&quot;/g, '"').replace(/\s+/g, " ").trim() : "";
+    results.push({ title, url: href, snippet });
+  }
+  return results;
+}
+async function webFetch(url, extractLinks) {
+  const controller = new AbortController();
+  const timeout = setTimeout(() => controller.abort(), 15e3);
+  try {
+    const response = await fetch(url, {
+      headers: {
+        "User-Agent": WEB_USER_AGENT,
+        "Accept": "text/html,application/xhtml+xml",
+        "Accept-Language": "nl,en;q=0.9"
+      },
+      redirect: "follow",
+      signal: controller.signal
+    });
+    if (!response.ok) throw new Error(`Fetch failed: HTTP ${response.status}`);
+    const contentType = response.headers.get("content-type") || "";
+    if (!contentType.includes("text/html") && !contentType.includes("text/plain") && !contentType.includes("application/xhtml")) {
+      throw new Error(`Unsupported content type: ${contentType}. Only HTML/text pages supported.`);
+    }
+    const html = await response.text();
+    const links = [];
+    if (extractLinks) {
+      const linkRegex = /<a\s[^>]*href="([^"]*)"[^>]*>([\s\S]*?)<\/a>/gi;
+      let match;
+      while ((match = linkRegex.exec(html)) !== null) {
+        const href = match[1];
+        const text2 = match[2].replace(/<[^>]+>/g, "").trim();
+        if (href && href.startsWith("http")) {
+          links.push({ href, text: text2.slice(0, 100) });
+        }
+      }
+    }
+    let text = html.replace(/<script[\s\S]*?<\/script>/gi, "").replace(/<style[\s\S]*?<\/style>/gi, "").replace(/<noscript[\s\S]*?<\/noscript>/gi, "").replace(/<nav[\s\S]*?<\/nav>/gi, " [NAV] ").replace(/<header[\s\S]*?<\/header>/gi, " [HEADER] ").replace(/<footer[\s\S]*?<\/footer>/gi, " [FOOTER] ").replace(/<(br|hr)\s*\/?>/gi, "\n").replace(/<\/(p|div|h[1-6]|li|tr|section|article)>/gi, "\n").replace(/<[^>]+>/g, " ").replace(/&nbsp;/g, " ").replace(/&amp;/g, "&").replace(/&lt;/g, "<").replace(/&gt;/g, ">").replace(/&#x27;/g, "'").replace(/&quot;/g, '"').replace(/[ \t]+/g, " ").replace(/\n\s*\n/g, "\n").trim();
+    const MAX_TEXT = 15e3;
+    if (text.length > MAX_TEXT) {
+      text = text.slice(0, MAX_TEXT) + "\n\n[... truncated at 15000 chars ...]";
+    }
+    return { text, links };
+  } finally {
+    clearTimeout(timeout);
+  }
+}
 function textSimilarity(a, b) {
   if (a === b) return 1;
   const norm = (s) => s.toLowerCase().replace(/\s+/g, " ").trim();
@@ -996,6 +1176,196 @@ ${output}` }]
       };
     }
     // -----------------------------------------------------------------
+    // Lead Generation tools
+    // -----------------------------------------------------------------
+    case "agent-check-lead-exists": {
+      const websiteUrl = sanitizeString(args2.website_url, 500).replace(/\/+$/, "");
+      const companyName = sanitizeString(args2.company_name, 500);
+      if (!websiteUrl && !companyName) {
+        throw new Error("At least one of website_url or company_name is required");
+      }
+      const checks = [];
+      if (websiteUrl) {
+        const { data: urlMatch } = await supabase2.from("lead").select("id, company_name, website_url").eq("website_url", websiteUrl).maybeSingle();
+        if (urlMatch) {
+          return {
+            content: [{
+              type: "text",
+              text: `DUPLICATE: Lead already exists (URL match). ID: ${urlMatch.id}, Company: ${urlMatch.company_name}. SKIP this company.`
+            }]
+          };
+        }
+        checks.push(`URL "${websiteUrl}" not found`);
+      }
+      if (companyName) {
+        const normalized = normalizeCompanyName(companyName);
+        if (normalized.length >= 3) {
+          const { data: nameMatches } = await supabase2.from("lead").select("id, company_name, website_url").ilike("company_name", `%${normalized.split(" ")[0]}%`).limit(20);
+          if (nameMatches) {
+            for (const match of nameMatches) {
+              const matchNorm = normalizeCompanyName(match.company_name);
+              if (textSimilarity(normalized, matchNorm) > 0.6) {
+                return {
+                  content: [{
+                    type: "text",
+                    text: `DUPLICATE: Similar company found. ID: ${match.id}, Name: "${match.company_name}" (${match.website_url}). SKIP this company.`
+                  }]
+                };
+              }
+            }
+          }
+        }
+        checks.push(`Name "${companyName}" has no similar matches`);
+      }
+      return {
+        content: [{
+          type: "text",
+          text: `NOT FOUND: No duplicate detected. ${checks.join(". ")}. Safe to proceed.`
+        }]
+      };
+    }
+    case "agent-save-lead": {
+      const companyName = sanitizeString(args2.company_name, 500);
+      const websiteUrl = sanitizeString(args2.website_url, 500).replace(/\/+$/, "");
+      if (!companyName) throw new Error("company_name is required");
+      if (!websiteUrl) throw new Error("website_url is required");
+      const { data: existing } = await supabase2.from("lead").select("id, company_name").eq("website_url", websiteUrl).maybeSingle();
+      if (existing) {
+        return {
+          content: [{
+            type: "text",
+            text: `Lead already exists (URL match). ID: ${existing.id}, Company: "${existing.company_name}". Skipped.`
+          }]
+        };
+      }
+      const normalized = normalizeCompanyName(companyName);
+      if (normalized.length >= 3) {
+        const { data: nameMatches } = await supabase2.from("lead").select("id, company_name").ilike("company_name", `%${normalized.split(" ")[0]}%`).limit(20);
+        if (nameMatches) {
+          for (const match of nameMatches) {
+            if (textSimilarity(normalized, normalizeCompanyName(match.company_name)) > 0.6) {
+              return {
+                content: [{
+                  type: "text",
+                  text: `Lead already exists (name match). ID: ${match.id}, Name: "${match.company_name}". Skipped.`
+                }]
+              };
+            }
+          }
+        }
+      }
+      const fitScore = args2.fit_score ? clamp(Number(args2.fit_score), 1, 10) : null;
+      const { data: inserted, error } = await supabase2.from("lead").insert({
+        company_name: companyName,
+        website_url: websiteUrl,
+        industry: args2.industry ? sanitizeString(args2.industry, 200) : null,
+        region: args2.region ? sanitizeString(args2.region, 200) : null,
+        description: args2.description ? sanitizeString(args2.description, 2e3) : null,
+        potential_fit: args2.potential_fit ? sanitizeString(args2.potential_fit, 2e3) : null,
+        fit_score: fitScore,
+        estimated_company_size: args2.estimated_company_size ? sanitizeString(args2.estimated_company_size, 50) : null,
+        kvk_number: args2.kvk_number ? sanitizeString(args2.kvk_number, 20) : null,
+        contact_name: args2.contact_name ? sanitizeString(args2.contact_name, 200) : null,
+        contact_role: args2.contact_role ? sanitizeString(args2.contact_role, 200) : null,
+        contact_email: args2.contact_email ? sanitizeString(args2.contact_email, 200) : null,
+        contact_phone: args2.contact_phone ? sanitizeString(args2.contact_phone, 50) : null,
+        contact_linkedin: args2.contact_linkedin ? sanitizeString(args2.contact_linkedin, 500) : null,
+        general_email: args2.general_email ? sanitizeString(args2.general_email, 200) : null,
+        general_phone: args2.general_phone ? sanitizeString(args2.general_phone, 50) : null,
+        source_url: args2.source_url ? sanitizeString(args2.source_url, 500) : null,
+        target_id: args2.target_id ? sanitizeString(args2.target_id, 100) : null,
+        status: "new"
+      }).select("id").single();
+      if (error) throw new Error(`Failed to save lead: ${error.message}`);
+      return {
+        content: [{
+          type: "text",
+          text: `Lead saved: "${companyName}" (${websiteUrl}). ID: ${inserted.id}. Fit score: ${fitScore ?? "N/A"}.`
+        }]
+      };
+    }
+    case "agent-save-email-draft": {
+      const leadId = sanitizeString(args2.lead_id, 100);
+      const subject = sanitizeString(args2.subject, 200);
+      let body = sanitizeString(args2.body, 5e3);
+      const tone = sanitizeString(args2.tone, 50) || "professional";
+      if (!leadId) throw new Error("lead_id is required");
+      if (!subject) throw new Error("subject is required");
+      if (!body) throw new Error("body is required");
+      body = body.replace(/[\u2013\u2014]/g, ",");
+      const { data: inserted, error } = await supabase2.from("lead_email_draft").insert({ lead_id: leadId, subject, body, tone, status: "draft" }).select("id").single();
+      if (error) throw new Error(`Failed to save email draft: ${error.message}`);
+      await supabase2.from("lead").update({ status: "email_drafted", updated_at: (/* @__PURE__ */ new Date()).toISOString() }).eq("id", leadId).eq("status", "new");
+      return {
+        content: [{
+          type: "text",
+          text: `Email draft saved for lead ${leadId}. Draft ID: ${inserted.id}. Subject: "${subject}".`
+        }]
+      };
+    }
+    case "agent-complete-target": {
+      const targetId = sanitizeString(args2.target_id, 100);
+      const resultsCount = clamp(Number(args2.results_count) || 0, 0, 9999);
+      if (!targetId) throw new Error("target_id is required");
+      const { error } = await supabase2.from("lead_generation_target").update({
+        status: "completed",
+        results_count: resultsCount,
+        completed_at: (/* @__PURE__ */ new Date()).toISOString()
+      }).eq("id", targetId);
+      if (error) throw new Error(`Failed to complete target: ${error.message}`);
+      return {
+        content: [{
+          type: "text",
+          text: `Target ${targetId} marked completed. ${resultsCount} leads found.`
+        }]
+      };
+    }
+    // -----------------------------------------------------------------
+    // Web Tools
+    // -----------------------------------------------------------------
+    case "web-search": {
+      const query = sanitizeString(args2.query, 500);
+      if (!query) throw new Error("query is required");
+      const maxResults = clamp(Number(args2.max_results) || 15, 1, 30);
+      const results = await webSearch(query, maxResults);
+      if (results.length === 0) {
+        return { content: [{ type: "text", text: `No search results found for: "${query}"` }] };
+      }
+      const formatted = results.map(
+        (r, i) => `${i + 1}. ${r.title}
+   URL: ${r.url}
+   ${r.snippet}`
+      ).join("\n\n");
+      return {
+        content: [{
+          type: "text",
+          text: `Search results for "${query}" (${results.length} results):
+
+${formatted}`
+        }]
+      };
+    }
+    case "web-fetch": {
+      const url = sanitizeString(args2.url, 2e3);
+      if (!url) throw new Error("url is required");
+      if (!/^https?:\/\//i.test(url)) throw new Error("url must start with http:// or https://");
+      const extractLinks = Boolean(args2.extract_links);
+      const result = await webFetch(url, extractLinks);
+      let text = `Content from ${url} (${result.text.length} chars):
+
+${result.text}`;
+      if (extractLinks && result.links.length > 0) {
+        const linkList = result.links.slice(0, 50).map(
+          (l) => `  ${l.text ? l.text + ": " : ""}${l.href}`
+        ).join("\n");
+        text += `
+
+--- Links (${result.links.length} found, showing max 50) ---
+${linkList}`;
+      }
+      return { content: [{ type: "text", text }] };
+    }
+    // -----------------------------------------------------------------
     default:
       return { content: [{ type: "text", text: `Unknown agent tool: ${name}` }] };
   }
@@ -2787,7 +3157,13 @@ async function main() {
       "/api/save-documentation": "agent-save-documentation",
       "/api/list-findings": "agent-list-findings",
       "/api/get-documentation": "agent-get-documentation",
-      "/api/validate-suggestions": "agent-validate-suggestions"
+      "/api/validate-suggestions": "agent-validate-suggestions",
+      "/api/check-lead-exists": "agent-check-lead-exists",
+      "/api/save-lead": "agent-save-lead",
+      "/api/save-email-draft": "agent-save-email-draft",
+      "/api/complete-target": "agent-complete-target",
+      "/api/web-search": "web-search",
+      "/api/web-fetch": "web-fetch"
     };
     const httpServer = createServer(async (req, res) => {
       const url = new URL(req.url ?? "/", `http://localhost:${httpPort}`);