@agent-pattern-labs/leads-rig 0.1.5 → 0.1.6

package/.claude/agents/general-free.md

@@ -27,7 +27,9 @@ Skip this only if the task explicitly says to attach to an existing session.
 ## Do
 
 - Crawl only public, high-signal company pages.
-- Extract emails, role inboxes, contact forms, public names/titles, source URL, evidence, and warnings.
+- Use your judgment on each email candidate: only emit an email lead when public evidence identifies one specific human owner of the address.
+- Extract high-quality named human emails, contact forms, public names/titles, source URL, evidence, and warnings.
+- If an address is shared, role-based, departmental, generic, blocked, or not tied to a visible named human, do not emit it as an email lead; emit a `contact_path` if useful.
 - Emit JSON/JSONL matching `templates/lead-schema.json`.
 - Run `npx public-leads validate --input <file>` before returning a result.
 - Return file paths and terminal status, not a prose-only summary.

package/.cursor/rules/agent-general-free.mdc

@@ -26,7 +26,9 @@ Skip this only if the task explicitly says to attach to an existing session.
 ## Do
 
 - Crawl only public, high-signal company pages.
-- Extract emails, role inboxes, contact forms, public names/titles, source URL, evidence, and warnings.
+- Use your judgment on each email candidate: only emit an email lead when public evidence identifies one specific human owner of the address.
+- Extract high-quality named human emails, contact forms, public names/titles, source URL, evidence, and warnings.
+- If an address is shared, role-based, departmental, generic, blocked, or not tied to a visible named human, do not emit it as an email lead; emit a `contact_path` if useful.
 - Emit JSON/JSONL matching `templates/lead-schema.json`.
 - Run `npx public-leads validate --input <file>` before returning a result.
 - Return file paths and terminal status, not a prose-only summary.

package/.cursor/rules/main.mdc

@@ -10,6 +10,7 @@ Agentic public-web lead discovery harness. It crawls company domains, extracts r
 ## Hard Limits
 
 - [H1] Only collect public-source business contact data. Do not infer, guess, brute-force, or enrich private email addresses. Every lead needs a source URL and evidence excerpt. The primary goal is named people, not generic company inboxes.
+- [H8] Apply coding-agent judgment to every email candidate. An email may pass only when public source evidence identifies one specific human owner of that email. Shared, role, departmental, generic, blocked, unknown-owner, or unnamed person-like emails must not be emitted as email leads.
 - [H2] Respect robots.txt, terms, rate limits, and site intent. If a page blocks crawling or is not public, record a warning and move on.
 - [H3] Never send outreach from this harness. The output is reviewable lead records and ingest-ready artifacts, not email campaigns.
 - [H4] Use file-backed artifacts before ingest. A subagent prose summary is not enough; write validated JSON/JSONL first, then ingest.
@@ -23,7 +24,7 @@ Agentic public-web lead discovery harness. It crawls company domains, extracts r
 - [D2] Use deterministic local helpers when they can validate, normalize, manifest, preflight, postflight, or ingest artifacts.
 - [D3] Read the active mode file before executing. Mode files own output shapes and gates.
 - [D4] Prefer official company websites and high-signal pages: home, contact, about, team, leadership, people, press, blog, careers, and legal pages.
-- [D5] Keep artifacts source-backed. Every emitted lead should include `domain`, `sourceUrl`, `evidence`, `emailType`, `extractionMethod`, `verificationStatus`, and `confidence`. Reject generic catch-all inboxes such as `info@`, `hello@`, `contact@`, `support@`, `team@`, and similar aliases; if no named person is available, prefer a `contact_path` record instead.
+- [D5] Keep artifacts source-backed. Every emitted lead should include `domain`, `sourceUrl`, `evidence`, `emailType`, `extractionMethod`, `verificationStatus`, and `confidence`. Reject generic catch-all inboxes such as `info@`, `hello@`, `contact@`, `support@`, `team@`, and similar aliases; if no named human email is available, prefer a `contact_path` record instead.
 
 ## Procedure
 

package/.opencode/agents/general-free.md

@@ -36,7 +36,9 @@ Skip this only if the task explicitly says to attach to an existing session.
 ## Do
 
 - Crawl only public, high-signal company pages.
-- Extract emails, role inboxes, contact forms, public names/titles, source URL, evidence, and warnings.
+- Use your judgment on each email candidate: only emit an email lead when public evidence identifies one specific human owner of the address.
+- Extract high-quality named human emails, contact forms, public names/titles, source URL, evidence, and warnings.
+- If an address is shared, role-based, departmental, generic, blocked, or not tied to a visible named human, do not emit it as an email lead; emit a `contact_path` if useful.
 - Emit JSON/JSONL matching `templates/lead-schema.json`.
 - Run `npx public-leads validate --input <file>` before returning a result.
 - Return file paths and terminal status, not a prose-only summary.

package/.pi/skills/general-free/SKILL.md

@@ -26,7 +26,9 @@ Skip this only if the task explicitly says to attach to an existing session.
 ## Do
 
 - Crawl only public, high-signal company pages.
-- Extract emails, role inboxes, contact forms, public names/titles, source URL, evidence, and warnings.
+- Use your judgment on each email candidate: only emit an email lead when public evidence identifies one specific human owner of the address.
+- Extract high-quality named human emails, contact forms, public names/titles, source URL, evidence, and warnings.
+- If an address is shared, role-based, departmental, generic, blocked, or not tied to a visible named human, do not emit it as an email lead; emit a `contact_path` if useful.
 - Emit JSON/JSONL matching `templates/lead-schema.json`.
 - Run `npx public-leads validate --input <file>` before returning a result.
 - Return file paths and terminal status, not a prose-only summary.

package/AGENTS.md

@@ -5,6 +5,7 @@ Agentic public-web lead discovery harness. It crawls company domains, extracts r
 ## Hard Limits
 
 - [H1] Only collect public-source business contact data. Do not infer, guess, brute-force, or enrich private email addresses. Every lead needs a source URL and evidence excerpt. The primary goal is named people, not generic company inboxes.
+- [H8] Apply coding-agent judgment to every email candidate. An email may pass only when public source evidence identifies one specific human owner of that email. Shared, role, departmental, generic, blocked, unknown-owner, or unnamed person-like emails must not be emitted as email leads.
 - [H2] Respect robots.txt, terms, rate limits, and site intent. If a page blocks crawling or is not public, record a warning and move on.
 - [H3] Never send outreach from this harness. The output is reviewable lead records and ingest-ready artifacts, not email campaigns.
 - [H4] Use file-backed artifacts before ingest. A subagent prose summary is not enough; write validated JSON/JSONL first, then ingest.
@@ -18,7 +19,7 @@ Agentic public-web lead discovery harness. It crawls company domains, extracts r
 - [D2] Use deterministic local helpers when they can validate, normalize, manifest, preflight, postflight, or ingest artifacts.
 - [D3] Read the active mode file before executing. Mode files own output shapes and gates.
 - [D4] Prefer official company websites and high-signal pages: home, contact, about, team, leadership, people, press, blog, careers, and legal pages.
-- [D5] Keep artifacts source-backed. Every emitted lead should include `domain`, `sourceUrl`, `evidence`, `emailType`, `extractionMethod`, `verificationStatus`, and `confidence`. Reject generic catch-all inboxes such as `info@`, `hello@`, `contact@`, `support@`, `team@`, and similar aliases; if no named person is available, prefer a `contact_path` record instead.
+- [D5] Keep artifacts source-backed. Every emitted lead should include `domain`, `sourceUrl`, `evidence`, `emailType`, `extractionMethod`, `verificationStatus`, and `confidence`. Reject generic catch-all inboxes such as `info@`, `hello@`, `contact@`, `support@`, `team@`, and similar aliases; if no named human email is available, prefer a `contact_path` record instead.
 
 ## Procedure
 

package/CLAUDE.md

@@ -5,6 +5,7 @@ Agentic public-web lead discovery harness. It crawls company domains, extracts r
 ## Hard Limits
 
 - [H1] Only collect public-source business contact data. Do not infer, guess, brute-force, or enrich private email addresses. Every lead needs a source URL and evidence excerpt. The primary goal is named people, not generic company inboxes.
+- [H8] Apply coding-agent judgment to every email candidate. An email may pass only when public source evidence identifies one specific human owner of that email. Shared, role, departmental, generic, blocked, unknown-owner, or unnamed person-like emails must not be emitted as email leads.
 - [H2] Respect robots.txt, terms, rate limits, and site intent. If a page blocks crawling or is not public, record a warning and move on.
 - [H3] Never send outreach from this harness. The output is reviewable lead records and ingest-ready artifacts, not email campaigns.
 - [H4] Use file-backed artifacts before ingest. A subagent prose summary is not enough; write validated JSON/JSONL first, then ingest.
@@ -18,7 +19,7 @@ Agentic public-web lead discovery harness. It crawls company domains, extracts r
 - [D2] Use deterministic local helpers when they can validate, normalize, manifest, preflight, postflight, or ingest artifacts.
 - [D3] Read the active mode file before executing. Mode files own output shapes and gates.
 - [D4] Prefer official company websites and high-signal pages: home, contact, about, team, leadership, people, press, blog, careers, and legal pages.
-- [D5] Keep artifacts source-backed. Every emitted lead should include `domain`, `sourceUrl`, `evidence`, `emailType`, `extractionMethod`, `verificationStatus`, and `confidence`. Reject generic catch-all inboxes such as `info@`, `hello@`, `contact@`, `support@`, `team@`, and similar aliases; if no named person is available, prefer a `contact_path` record instead.
+- [D5] Keep artifacts source-backed. Every emitted lead should include `domain`, `sourceUrl`, `evidence`, `emailType`, `extractionMethod`, `verificationStatus`, and `confidence`. Reject generic catch-all inboxes such as `info@`, `hello@`, `contact@`, `support@`, `team@`, and similar aliases; if no named human email is available, prefer a `contact_path` record instead.
 
 ## Procedure
 

package/README.md

@@ -12,6 +12,7 @@ Agentic public-web lead discovery harness for portable, source-backed lead artif
 - Ships OpenCode, Claude Code, Cursor, Codex, and Pi harness config from one `iso/` source.
 
 The harness is optimized to find named people who can actually be contacted. When only a general company contact path exists, it should emit a `contact_path` record rather than a generic organizational inbox.
+CLI summaries distinguish total reviewable records from `goodLeads`; only non-generic `person` email records with a named human owner visible in the evidence count as good leads. Role, shared, departmental, unknown-owner, blocked, unsupported, or unnamed person-like emails are rejected by validation.
 
 ## Core Commands
 
@@ -70,7 +71,7 @@ gh run watch
 You can also publish by creating a GitHub release for the package version:
 
 ```bash
-gh release create v0.1.4 --title v0.1.4 --generate-notes
+gh release create v0.1.6 --title v0.1.6 --generate-notes
 ```
 
 The package version must not already exist on npm.

package/bin/create-leads-harness.mjs

@@ -51,7 +51,7 @@ write('package.json', JSON.stringify({
     'update-harness': 'npm install @agent-pattern-labs/leads-rig@latest && public-leads sync',
   },
   dependencies: {
-    '@agent-pattern-labs/leads-rig': '^0.1.4',
+    '@agent-pattern-labs/leads-rig': '^0.1.6',
   },
   engines: {
     node: '>=20.6.0',

package/docs/ARCHITECTURE.md

@@ -82,4 +82,4 @@ The package also carries the `@agent-pattern-labs/iso-*` helper ecosystem for tr
 - page visits
 - ingest requests
 
-The validator intentionally accepts the same defaults the local runtime normalizes, but it fails missing source evidence, invalid URL fields, invalid `emailType`, invalid `verificationStatus`, invalid confidence, person/role leads without email, and generic catch-all inboxes such as `info@`, `hello@`, or similar organizational aliases.
+The validator intentionally accepts the same defaults the local runtime normalizes, but it fails missing source evidence, invalid URL fields, invalid `emailType`, invalid `verificationStatus`, invalid confidence, person leads without email, and generic catch-all inboxes such as `info@`, `hello@`, or similar organizational aliases. Any email record must be a named `person` lead with a non-generic email and a human owner visible in the evidence; role inboxes, unknown-owner emails, blocked emails, unsupported emails, and unnamed person-like emails fail validation. Summaries expose `goodLeadCount`, which only counts those high-quality named human email records.

package/docs/CONSTRUCTION.md

@@ -37,4 +37,4 @@
 6. Run `npm audit`.
 7. Run `npm pack` and confirm generated runtime surfaces are included.
 8. Scaffold a clean consumer project and test install-time sync from the packed package or source tree.
-9. Publish through GitHub Actions with `gh workflow run publish.yml --ref main`, or create a published release with `gh release create v0.1.4 --title v0.1.4 --generate-notes`.
+9. Publish through GitHub Actions with `gh workflow run publish.yml --ref main`, or create a published release with `gh release create v0.1.6 --title v0.1.6 --generate-notes`.

package/iso/agents/general-free.md

@@ -38,7 +38,9 @@ Skip this only if the task explicitly says to attach to an existing session.
 ## Do
 
 - Crawl only public, high-signal company pages.
-- Extract emails, role inboxes, contact forms, public names/titles, source URL, evidence, and warnings.
+- Use your judgment on each email candidate: only emit an email lead when public evidence identifies one specific human owner of the address.
+- Extract high-quality named human emails, contact forms, public names/titles, source URL, evidence, and warnings.
+- If an address is shared, role-based, departmental, generic, blocked, or not tied to a visible named human, do not emit it as an email lead; emit a `contact_path` if useful.
 - Emit JSON/JSONL matching `templates/lead-schema.json`.
 - Run `npx public-leads validate --input <file>` before returning a result.
 - Return file paths and terminal status, not a prose-only summary.

package/iso/instructions.md

@@ -5,6 +5,7 @@ Agentic public-web lead discovery harness. It crawls company domains, extracts r
 ## Hard Limits
 
 - [H1] Only collect public-source business contact data. Do not infer, guess, brute-force, or enrich private email addresses. Every lead needs a source URL and evidence excerpt. The primary goal is named people, not generic company inboxes.
+- [H8] Apply coding-agent judgment to every email candidate. An email may pass only when public source evidence identifies one specific human owner of that email. Shared, role, departmental, generic, blocked, unknown-owner, or unnamed person-like emails must not be emitted as email leads.
 - [H2] Respect robots.txt, terms, rate limits, and site intent. If a page blocks crawling or is not public, record a warning and move on.
 - [H3] Never send outreach from this harness. The output is reviewable lead records and ingest-ready artifacts, not email campaigns.
 - [H4] Use file-backed artifacts before ingest. A subagent prose summary is not enough; write validated JSON/JSONL first, then ingest.
@@ -18,7 +19,7 @@ Agentic public-web lead discovery harness. It crawls company domains, extracts r
 - [D2] Use deterministic local helpers when they can validate, normalize, manifest, preflight, postflight, or ingest artifacts.
 - [D3] Read the active mode file before executing. Mode files own output shapes and gates.
 - [D4] Prefer official company websites and high-signal pages: home, contact, about, team, leadership, people, press, blog, careers, and legal pages.
-- [D5] Keep artifacts source-backed. Every emitted lead should include `domain`, `sourceUrl`, `evidence`, `emailType`, `extractionMethod`, `verificationStatus`, and `confidence`. Reject generic catch-all inboxes such as `info@`, `hello@`, `contact@`, `support@`, `team@`, and similar aliases; if no named person is available, prefer a `contact_path` record instead.
+- [D5] Keep artifacts source-backed. Every emitted lead should include `domain`, `sourceUrl`, `evidence`, `emailType`, `extractionMethod`, `verificationStatus`, and `confidence`. Reject generic catch-all inboxes such as `info@`, `hello@`, `contact@`, `support@`, `team@`, and similar aliases; if no named human email is available, prefer a `contact_path` record instead.
 
 ## Procedure
 

package/lib/leadharness-crawler.mjs

@@ -1,7 +1,7 @@
 import { createHash } from 'crypto';
 import { lookup, resolveMx } from 'dns/promises';
 import { setTimeout as sleep } from 'timers/promises';
-import { cleanDomains, dedupeLeadRecords, isRejectedLeadEmail, mergeLeadRecords, normalizeDomain } from './leadharness-leads.mjs';
+import { cleanDomains, dedupeLeadRecords, isGoodLead, isRejectedLeadEmail, mergeLeadRecords, normalizeDomain } from './leadharness-leads.mjs';
 
 const DEFAULT_MAX_PAGES = 10;
 const HARD_MAX_PAGES = 25;
@@ -575,6 +575,7 @@ function classifyEmailLocal(local) {
 }
 
 function isRelevantLead(lead, options) {
+  if (lead.email && !isGoodLead(lead)) return false;
   if (!options.includeBlocked && lead.emailType === 'blocked') return false;
   return lead.confidence >= options.minConfidence;
 }

package/lib/leadharness-leads.mjs

@@ -296,6 +296,9 @@ export function validateLead(lead, path, issues) {
   if (lead.email && isRejectedLeadEmail(lead.email)) {
     issues.push(issue('error', `${path}.email`, 'generic_inbox_rejected', 'generic company inboxes such as info@, hello@, or support@ are not accepted; submit named people instead'));
   }
+  if (lead.email && !isRejectedLeadEmail(lead.email) && ALLOWED_EMAIL_TYPES.has(lead.emailType) && !isGoodLead(lead)) {
+    issues.push(issue('error', `${path}.emailType`, 'non_person_email_rejected', 'email records must identify one named human visible in the evidence; role, unknown, blocked, unsupported, or unnamed person emails are not accepted'));
+  }
   if (lead.emailType && !ALLOWED_EMAIL_TYPES.has(lead.emailType)) {
     issues.push(issue('error', `${path}.emailType`, 'invalid_email_type', `emailType must be one of ${[...ALLOWED_EMAIL_TYPES].join(', ')}`));
   }
@@ -381,11 +384,13 @@ export function summarizePayload(payload) {
   const byType = {};
   let confidenceTotal = 0;
   let withEmail = 0;
+  let goodLeadCount = 0;
   const uniqueEmails = new Set();
   const mxVerifiedEmails = new Set();
   for (const lead of leads) {
     byType[lead.emailType || 'unknown'] = (byType[lead.emailType || 'unknown'] || 0) + 1;
     confidenceTotal += Number.isFinite(lead.confidence) ? lead.confidence : 0;
+    if (isGoodLead(lead)) goodLeadCount++;
     if (lead.email) {
       withEmail++;
       uniqueEmails.add(String(lead.email).toLowerCase());
@@ -399,6 +404,8 @@ export function summarizePayload(payload) {
     domainCount: domains.length,
     leadCount: leads.length,
     resultCount: results.length,
+    goodLeadCount,
+    reviewOnlyLeadCount: Math.max(0, leads.length - goodLeadCount),
     withEmail,
     uniqueEmailCount: uniqueEmails.size,
     mxVerifiedEmailCount: mxVerifiedEmails.size,
@@ -417,6 +424,7 @@ export function createManifestRecord({ inputPath, payload, validation, generated
     generatedAt,
     domains: payload.domains,
     leadCount: payload.leads.length,
+    goodLeadCount: validation.summary?.goodLeadCount || 0,
     resultCount: payload.results.length,
     errorCount: payload.errors.length,
     validation: {
@@ -648,7 +656,10 @@ function stableLeadKey(lead) {
 }
 
 function leadQualityScore(lead) {
-  return (lead?.confidence || 0) * 1000
+  const goodLead = isGoodLead(lead);
+  return (goodLead ? 50_000 : 0)
+    + (lead?.email && !goodLead ? -20_000 : 0)
+    + (lead?.confidence || 0) * 1000
     + verificationRank(lead?.verificationStatus) * 100
     + pagePriority(lead?.sourceUrl || '') * 10
     + (Array.isArray(lead?.sources) ? lead.sources.length : 0) * 5
@@ -738,6 +749,30 @@ export function isRejectedLeadLocalPart(local) {
   });
 }
 
+export function isGoodLead(lead) {
+  if (!lead || typeof lead !== 'object') return false;
+  if (stringValue(lead.emailType).toLowerCase() !== 'person') return false;
+  if (!lead.email || isRejectedLeadEmail(lead.email)) return false;
+  const contactName = stringValue(lead.contactName);
+  if (!contactName) return false;
+  if (!evidenceIdentifiesContact(lead, contactName)) return false;
+  return true;
+}
+
+function evidenceIdentifiesContact(lead, contactName) {
+  const needle = compactComparable(contactName);
+  if (!needle) return false;
+  const evidence = [
+    lead.evidence,
+    ...(Array.isArray(lead.sources) ? lead.sources.map((source) => source?.evidence) : []),
+  ].map(compactComparable).join(' ');
+  return evidence.includes(needle);
+}
+
+function compactComparable(value) {
+  return stringValue(value).toLowerCase().replace(/[^a-z0-9]+/g, ' ').trim();
+}
+
 function normalizeRejectedLeadLocalPart(local) {
   let value = String(local || '').trim().toLowerCase();
   for (const noise of ['mailto', 'u003e', 'email']) {

package/modes/_shared.md

@@ -28,18 +28,21 @@ Required fields:
 - Prioritize home, contact, about, team, leadership, people, press, blog, careers, and legal pages.
 - Respect robots.txt, paywalls, login walls, and obvious anti-scraping notices.
 - Do not infer email formats. Do not brute-force guessed addresses.
+- Use coding-agent judgment for every email candidate. Only emit an email lead when the source evidence identifies one specific human behind that address.
 - Prefer named people over organizational inboxes. Do not emit generic catch-all emails such as `info@`, `hello@`, `contact@`, `support@`, `team@`, or similar aliases. If only a general contact path exists, emit `contact_path` instead.
+- Reject email records that are role-based, shared, departmental, unknown ownership, blocked operational inboxes, or person-like but missing a visible human name in the source evidence.
 - Keep contact forms as `emailType: "contact_path"` with empty `email`.
 - Mark operational no-contact inboxes such as `noreply`, `abuse`, `security`, and `legal` as `blocked` with confidence `0`.
 
 ## Quality Heuristics
 
-- `person`: named person or person-like email with source context.
-- `role`: public role inbox only when it is clearly role-specific and not a generic catch-all alias.
+- `goodLeadCount`: counts only `person` email records with a non-generic email and named `contactName` visible in the evidence. Generic, role, unknown, blocked, unnamed, unsupported, and `contact_path` records are not good leads.
+- `person`: named human with source context tying that person to the email address.
+- `role`: legacy/review-only classification. Do not emit role email records in new artifacts.
 - `contact_path`: public form or page where no email is available.
 - `blocked`: public address that should not be used for outbound.
 
-Confidence should reflect source quality, person specificity, domain match, and verification. Low-confidence records are allowed if they preserve warnings and evidence.
+Confidence should reflect source quality, person specificity, domain match, and verification. Low-confidence contact paths are allowed if they preserve warnings and evidence; email leads should be high-quality named humans only.
 
 ## Local Commands
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@agent-pattern-labs/leads-rig",
-  "version": "0.1.5",
+  "version": "0.1.6",
   "description": "Agentic public-web lead discovery harness with portable ingest artifacts",
   "type": "module",
   "author": "Razroo",

package/scripts/crawl.mjs

@@ -57,7 +57,7 @@ async function main() {
     console.log(JSON.stringify({ output: out, validation, summary: validation.summary }, null, 2));
   } else {
     console.log(`crawl: wrote ${relativeProjectPath(resolveProjectPath(out))}`);
-    console.log(`domains=${validation.summary.domainCount} leads=${validation.summary.leadCount} results=${validation.summary.resultCount} errors=${payload.errors.length}`);
+    console.log(`domains=${validation.summary.domainCount} leads=${validation.summary.leadCount} goodLeads=${validation.summary.goodLeadCount} results=${validation.summary.resultCount} errors=${payload.errors.length}`);
   }
 
   if (!validation.ok) {

package/scripts/manifest.mjs

@@ -50,6 +50,7 @@ try {
     batchCount: manifest.batches.length,
     readyForIngest: manifest.batches.filter((item) => item.readyForIngest).length,
     leadCount: manifest.batches.reduce((sum, item) => sum + item.leadCount, 0),
+    goodLeadCount: manifest.batches.reduce((sum, item) => sum + (item.goodLeadCount || 0), 0),
     errorCount: manifest.batches.reduce((sum, item) => sum + item.validation.errors, 0),
     warningCount: manifest.batches.reduce((sum, item) => sum + item.validation.warnings, 0),
   };
@@ -60,8 +61,8 @@ try {
     console.log(JSON.stringify({ manifest: relativeProjectPath(manifestPath), record, summary: manifest.summary }, null, 2));
   } else {
     console.log(`manifest: ${relativeProjectPath(manifestPath)}`);
-    console.log(`batch: ${record.id} leads=${record.leadCount} ready=${record.readyForIngest}`);
-    console.log(`summary: batches=${manifest.summary.batchCount} ready=${manifest.summary.readyForIngest} leads=${manifest.summary.leadCount}`);
+    console.log(`batch: ${record.id} leads=${record.leadCount} goodLeads=${record.goodLeadCount} ready=${record.readyForIngest}`);
+    console.log(`summary: batches=${manifest.summary.batchCount} ready=${manifest.summary.readyForIngest} leads=${manifest.summary.leadCount} goodLeads=${manifest.summary.goodLeadCount}`);
   }
 
   process.exit(validation.ok ? 0 : 1);

package/scripts/pipeline.mjs

@@ -66,7 +66,7 @@ try {
   const manifest = upsertManifest(manifestPath, out, payload, validation);
   console.log(`pipeline: wrote ${relativeProjectPath(resolveProjectPath(out))}`);
   console.log(`manifest: ${relativeProjectPath(resolveProjectPath(manifestPath))}`);
-  console.log(`domains=${validation.summary.domainCount} leads=${validation.summary.leadCount} results=${validation.summary.resultCount} errors=${payload.errors.length}`);
+  console.log(`domains=${validation.summary.domainCount} leads=${validation.summary.leadCount} goodLeads=${validation.summary.goodLeadCount} results=${validation.summary.resultCount} errors=${payload.errors.length}`);
 
   if (opts.ingest || opts.upload) {
     const ingestResult = await ingestPayload(payload, {
@@ -110,6 +110,7 @@ function upsertManifest(manifestPath, inputPath, payload, validation) {
     batchCount: manifest.batches.length,
     readyForIngest: manifest.batches.filter((item) => item.readyForIngest).length,
     leadCount: manifest.batches.reduce((sum, item) => sum + item.leadCount, 0),
+    goodLeadCount: manifest.batches.reduce((sum, item) => sum + (item.goodLeadCount || 0), 0),
     errorCount: manifest.batches.reduce((sum, item) => sum + item.validation.errors, 0),
     warningCount: manifest.batches.reduce((sum, item) => sum + item.validation.warnings, 0),
   };

package/scripts/validate-leads.mjs

@@ -58,7 +58,7 @@ try {
 function printReport(report) {
   const rel = relativeProjectPath(report.input);
   console.log(`${report.ok ? 'OK' : 'ERROR'} ${rel}`);
-  console.log(`domains=${report.summary.domainCount} leads=${report.summary.leadCount} results=${report.summary.resultCount} avgConfidence=${report.summary.averageConfidence}`);
+  console.log(`domains=${report.summary.domainCount} leads=${report.summary.leadCount} goodLeads=${report.summary.goodLeadCount} results=${report.summary.resultCount} avgConfidence=${report.summary.averageConfidence}`);
   if (Object.keys(report.summary.byType).length > 0) {
     console.log(`types=${Object.entries(report.summary.byType).map(([type, count]) => `${type}:${count}`).join(', ')}`);
   }