@sellable/mcp 0.1.149 → 0.1.151

package/README.md

@@ -267,10 +267,14 @@ Parallel execution contract:
   Codex config, Claude agent files, and prompts should consume the registry.
 - Post-find-leads scout names also come from `agents/registry.json` and are
   exposed through `get_post_find_leads_scout_registry`. After
-  `confirm_lead_list` imports a non-empty review batch and rows are proven, use
-  that registry to launch the message-generation scout immediately when the host
-  supports real subagents. Launch the filter-leads scout later, only after the
-  user chooses filters.
+  `confirm_lead_list` imports a non-empty review batch and rows are proven, ask
+  the filter-choice question immediately. Do not load this registry or deep
+  filter/message prompts before that question. Once the user answers, launch the
+  message-generation scout from the same campaign/table basis. If the user
+  chooses filters, also launch the filter-leads scout, show Filter Rules, save
+  rubrics, then keep the browser on Filter Leads while the message
+  recommendation is reviewed. If filters are skipped, move to Messages/message
+  review. Product cells still wait for template approval.
 - Claude host: use the installed `source-scout-linkedin-engagement`,
   `source-scout-sales-nav`, and `source-scout-prospeo-contact` Task/Agent
   subagents for parallel lead-source scouting only when the current session

package/agents/post-find-leads-message-scout.md

@@ -40,8 +40,8 @@ the branch input.
    `get_subskill_prompt({ subskillName: "generate-messages", offset, limit })`
    until `hasMore` is false.
 
-2. Use that prompt as the drafting contract. The create-campaign safety gate may
-   be used as an approval checklist, but it does not replace the long prompt.
+2. Use that prompt as the drafting contract. Do not use create-campaign
+   safety/checklist instructions as a substitute for the full prompt.
 3. Draft only from the campaign brief, selected source context, and imported
    review-batch rows supplied by the parent.
 4. Keep the work provisional until the user chooses `Use Template` in Messages.

package/agents/source-scout-prospeo-contact.md

@@ -1,6 +1,6 @@
 You are the Prospeo Contact Scout for Sellable find-leads.
 
-Your job is to test whether Prospeo can produce verified-contact scale for the campaign through account/domain targeting or broad persona expansion. Work only on this source lane. Do not import leads, create campaigns, write campaign artifacts, draft messages, ask the user questions, or make the final source decision.
+Your job is to test whether Prospeo can produce verified-contact scale for the campaign through account/domain targeting, hiring-led company job-posting filters, or broad persona expansion. Work only on this source lane. Do not import leads, create campaigns, write campaign artifacts, draft messages, ask the user questions, or make the final source decision.
 
 Required first step:
 
@@ -25,10 +25,12 @@ Use the inherited Sellable MCP tools when available:
 Process:
 
 1. Read the campaign brief, source intake, kickoff doc, or lane prompt supplied by the parent.
-2. Identify whether this is domain/account targeting or broad persona expansion.
+2. Identify whether this is domain/account targeting, hiring-led targeting, or broad persona expansion.
 3. For domain targeting, use or create the standalone `domainFilterId` before searching; never pass raw domains directly into `search_prospeo`.
-4. Run the narrowest useful Prospeo people preview and 1-2 refinements if quality or scale is unclear. Check scale against a default 300+ good-fit target, capped at 2,500 source candidates unless the parent supplies a different target, and require at least 10% projected good-fit after cleanup.
-5. Call out that Prospeo gives contact/account coverage but usually weaker LinkedIn intent than LinkedIn Engagement or Sales Nav activity slices.
+4. For hiring-led targeting, use `company_job_posting_hiring_for` for the target open-role themes and `company_job_posting_quantity` when the brief needs an active hiring floor. Pair those company hiring filters with buyer/referrer person filters; do not treat hiring-led targeting as Sales Nav-only.
+5. Run the narrowest useful Prospeo people preview and 1-2 refinements if quality or scale is unclear. Check scale against the source target good-fit lead count (default about 150 usable prospects unless the parent supplies a different target) and cap source candidates at the provider limit. Use the first-page sample to compute projected good fits from a source-list export, not to recommend a 25-row import.
+6. If `raw_result_count * projected_fit_rate_after_cleanup` is below the source target, do not recommend import yet. Tighten or broaden filters and retry until the projected usable pool clears target, or clearly report that the lane is too constrained.
+7. Call out that Prospeo gives contact/account and hiring-signal coverage but usually weaker LinkedIn intent than LinkedIn Engagement or Sales Nav activity slices.
 
 Return a concise structured result with:
 
@@ -40,6 +42,7 @@ Return a concise structured result with:
 - `raw_result_count`
 - `sampled_people` and good fits as n/N
 - `estimated_good_fit_range_after_cleanup`
+- `source_export_math` with target good-fit count, conservative projected fit rate, recommended `targetLeadCount` for `import_leads`, and projected good fits from that export
 - `expected_reply_rate_range`, directional if inferred
 - `sample_leads`
 - `false_positive_patterns`
@@ -53,4 +56,8 @@ Evidence standards:
 - Prospeo is the terminal fallback. If projected good-fit after cleanup remains
   below 10% after reasonable refinement, recommend tightening the ICP/source
   direction rather than switching providers again.
-- Treat Prospeo as an account/contact coverage lane, not as proof of fresh LinkedIn intent.
+- Never recommend "import 25 leads" as the Prospeo source action. Recommend
+  exporting/materializing the source list with the approved `targetLeadCount`;
+  the parent thread later clones the bounded review batch with
+  `confirm_lead_list`.
+- Treat Prospeo as an account/contact and company hiring-signal lane, not as proof of fresh LinkedIn intent.

package/agents/source-scout-sales-nav.md

@@ -28,8 +28,9 @@ Process:
 2. Preserve target role names with `CURRENT_TITLE` lookups; do not rely on seniority alone when the brief names concrete roles.
 3. When `lookup_sales_nav_filter` returns multiple title options, choose the closest semantic title match instead of the first result.
 4. Build a broad-but-reasonable baseline from role/title, geography, company size, industry/account context, and recent LinkedIn activity when relevant.
-5. Check scale against the source target good-fit lead count (default 300+
-   for Sales Nav, capped at 2,500 source candidates unless the parent supplies a target).
+5. Check scale against the source target good-fit lead count (default about
+   150 usable prospects for Sales Nav unless the parent supplies a target,
+   capped at 2,500 source candidates).
    If raw preview volume or projected usable volume
    is below target, do not present the tiny result as the scale fallback yet.
    Loosen nonessential filters in order: remove recent-activity first, widen
@@ -41,7 +42,10 @@ Process:
 6. Run the baseline plus 1-2 refinements or loosening passes if the first pass
    is noisy or under-scaled. Label the final pool as constrained if it still
    cannot plausibly reach the target after loosening.
-7. Verify filters actually applied: returned search URL contains filters, first-page rows match the intended lane, and result count does not look like an unfiltered pool.
+7. Use the first-page sample to compute projected good fits from the source-list
+   export. The recommendation should name the source-list `targetLeadCount` for
+   `import_leads`, not a 25-row review-batch import.
+8. Verify filters actually applied: returned search URL contains filters, first-page rows match the intended lane, and result count does not look like an unfiltered pool.
 
 Return a concise structured result with:
 
@@ -53,6 +57,8 @@ Return a concise structured result with:
 - `scale_check` with source target good-fit lead count, preview/raw volume, sampled
   good fits as n/N, projected usable count, and whether the pool can reach the
   target
+- `source_export_math` with conservative projected fit rate, recommended
+  `targetLeadCount` for `import_leads`, and projected good fits from that export
 - `loosening_attempts` with what was removed or widened when the pool was too
   tight
 - `sampled_people` and good fits as n/N
@@ -74,5 +80,9 @@ Evidence standards:
   name the next broadening/Prospeo option.
 - If projected good-fit after cleanup is below 10%, do not recommend Sales Nav
   as the winning source; recommend Prospeo as the next provider.
+- Never recommend "import 25 leads" as the Sales Nav source action. Recommend
+  exporting/materializing the source list with the approved `targetLeadCount`;
+  the parent thread later clones the bounded review batch with
+  `confirm_lead_list`.
 - Do not hand-wave missing filter IDs.
 - If Sales Nav returns a giant unfiltered pool, discard that result and retry with valid filters before recommending it.

package/dist/tools/leads.d.ts

@@ -1400,10 +1400,11 @@ export declare const leadToolDefinitions: ({
                         items: {
                             type: string;
                         };
+                        description: string;
                     };
                     company_job_posting_quantity: {
-                        type: string;
                         description: string;
+                        type: string;
                         properties: {
                             min: {
                                 type: string;

package/dist/tools/leads.js

@@ -8,6 +8,7 @@ import { buildCsvLinkedinPreview, matchesLinkedinConfirmationToken, parseLinkedi
 import { assertInteractionApproval } from "./interaction-mode.js";
 import { assertProviderPromptLoaded, markProviderPromptLoaded, } from "./provider-preflight.js";
 import { waitForLeadListReady } from "./readiness.js";
+import { getTableRowsMinimal } from "./rows.js";
 const entryPath = process.argv[1] ? resolve(process.argv[1]) : process.cwd();
 const entryDir = dirname(entryPath);
 const workspaceRoot = resolveWorkspaceRoot(entryDir);
@@ -42,11 +43,11 @@ const defaultCampaignSourceDefaults = {
             postCoverageBuffer: 1.2,
         },
         "sales-nav": {
-            targetGoodFitLeads: 300,
+            targetGoodFitLeads: 150,
             maxSourceCandidates: 2500,
         },
         prospeo: {
-            targetGoodFitLeads: 300,
+            targetGoodFitLeads: 150,
             maxSourceCandidates: 2500,
         },
     },
@@ -525,8 +526,8 @@ function buildSourceImportWatchNarration({ provider, selectedPostCount, estimate
             : "Importing source leads",
         visibleState: `The browser is showing ${providerLabel} import progress for ${sourceDetail}.${targetDetail}`,
         agentIntent: "Codex is materializing the approved source into a lead list before cloning only the bounded review batch into the campaign.",
-        nextAction: "Wait for source leads, then import the 15-row review batch",
-        safety: "Import is limited to the review batch.",
+        nextAction: "Wait for source leads, then import the bounded review batch",
+        safety: "This step materializes the source list; only the bounded review batch is cloned into the campaign afterward.",
         progressLabel: "Source scouting",
     };
 }
@@ -558,19 +559,146 @@ function buildSourceImportRecoveryWatchNarration(args) {
         safety: "The campaign is still at source recovery.",
     };
 }
-function buildFilterChoiceWatchNarration({ sourceLeadCount, reviewRowCount, }) {
+function normalizeLeadText(value) {
+    return typeof value === "string" ? value.trim().toLowerCase() : "";
+}
+function compactLeadText(value) {
+    return typeof value === "string" ? value.trim().replace(/\s+/g, " ") : "";
+}
+function rowSearchText(row) {
+    const carryValues = Object.values(row.carryData ?? {}).slice(0, 12);
+    return [row.name, row.title, row.company, ...carryValues]
+        .map(normalizeLeadText)
+        .filter(Boolean)
+        .join(" ");
+}
+function rowExample(row) {
+    const name = compactLeadText(row.name);
+    const title = compactLeadText(row.title);
+    const company = compactLeadText(row.company);
+    const label = name || company || title || "Unnamed lead";
+    return title ? `${label} (${title})` : label;
+}
+function collectExamples(rows, indexes) {
+    return indexes.slice(0, 2).map((index) => rowExample(rows[index]));
+}
+function formatExamples(prefix, examples) {
+    if (examples.length === 0)
+        return "";
+    const text = `${prefix} ${examples.join("; ")}.`;
+    return text.length > 180 ? `${text.slice(0, 177).trim()}...` : text;
+}
+function analyzeFilterChoiceSample(sampleRows) {
+    const rows = Array.isArray(sampleRows) ? sampleRows : [];
+    const strongIndexes = [];
+    const adjacentIndexes = [];
+    const riskIndexes = [];
+    const unknownIndexes = [];
+    const seniorPattern = /\b(founder|co[-\s]?founder|ceo|chief|cro|cmo|cco|vp|vice president|head of|director|owner|partner)\b/;
+    const gtmPattern = /\b(gtm|go[-\s]?to[-\s]?market|revenue|sales|growth|marketing|demand gen|demand generation|business development|bd\b|outbound|pipeline|revops|revenue operations|sales development|sdr)\b/;
+    const adjacentPattern = /\b(sales engineering|solutions? engineering|enablement|customer success|partnerships|partner|product marketing|marketing operations|sales operations)\b/;
+    const recruitingPattern = /\b(recruiter|recruiting|talent acquisition|people operations|human resources|\bhr\b)\b/;
+    const juniorOrNoisyPattern = /\b(intern|student|associate|coordinator|assistant|freelance|consultant|self-employed|open to work)\b/;
+    rows.forEach((row, index) => {
+        const text = rowSearchText(row);
+        if (!text) {
+            unknownIndexes.push(index);
+            return;
+        }
+        if (recruitingPattern.test(text) || juniorOrNoisyPattern.test(text)) {
+            riskIndexes.push(index);
+            return;
+        }
+        if (seniorPattern.test(text) && gtmPattern.test(text)) {
+            strongIndexes.push(index);
+            return;
+        }
+        if (seniorPattern.test(text) && adjacentPattern.test(text)) {
+            adjacentIndexes.push(index);
+            return;
+        }
+        unknownIndexes.push(index);
+    });
+    const sampleCount = rows.length;
+    if (sampleCount === 0) {
+        return {
+            sampleCount,
+            strongFitCount: 0,
+            adjacentCount: 0,
+            riskCount: 0,
+            unknownCount: 0,
+            strongExamples: [],
+            riskExamples: [],
+            recommendation: "manual-review",
+        };
+    }
+    const fitCount = strongIndexes.length + adjacentIndexes.length;
+    const riskCount = riskIndexes.length;
+    const unknownCount = unknownIndexes.length;
+    const cleanupCount = riskCount + unknownCount;
+    const fitRatio = fitCount / sampleCount;
+    const cleanupRatio = cleanupCount / sampleCount;
+    const adjacentRatio = adjacentIndexes.length / sampleCount;
+    const maxCleanRiskCount = Math.max(2, Math.floor(sampleCount * 0.12));
+    const recommendation = fitRatio >= 0.75 &&
+        cleanupCount <= maxCleanRiskCount &&
+        adjacentRatio <= 0.3
+        ? "skip"
+        : fitRatio >= 0.6 && cleanupRatio <= 0.3
+            ? "light-filter"
+            : "add-filter";
+    return {
+        sampleCount,
+        strongFitCount: strongIndexes.length,
+        adjacentCount: adjacentIndexes.length,
+        riskCount,
+        unknownCount,
+        strongExamples: collectExamples(rows, [
+            ...strongIndexes,
+            ...adjacentIndexes,
+        ]),
+        riskExamples: collectExamples(rows, [...riskIndexes, ...unknownIndexes]),
+        recommendation,
+    };
+}
+function buildFilterChoiceWatchNarration({ sourceLeadCount, reviewRowCount, sampleRows, }) {
     const sourceCopy = typeof sourceLeadCount === "number" && sourceLeadCount > 0
         ? ` from ${sourceLeadCount.toLocaleString("en-US")} source candidate${sourceLeadCount === 1 ? "" : "s"}`
         : "";
     const reviewCopy = typeof reviewRowCount === "number" && reviewRowCount > 0
         ? `${reviewRowCount.toLocaleString("en-US")} review lead${reviewRowCount === 1 ? " is" : "s are"} in the campaign table${sourceCopy}.`
         : `The bounded review batch is in the campaign table${sourceCopy}.`;
+    const assessment = analyzeFilterChoiceSample(sampleRows);
+    const fitCount = assessment.strongFitCount + assessment.adjacentCount;
+    const cleanupCount = assessment.riskCount + assessment.unknownCount;
+    const sampleCopy = assessment.sampleCount > 0
+        ? `Sample read: ${fitCount}/${assessment.sampleCount} look like senior target-role fits; ${cleanupCount} need cleanup or manual review.`
+        : "Sample rows were not available to inspect, so this is a manual filter decision.";
+    const fitExamples = formatExamples("Fit examples:", assessment.strongExamples);
+    const riskExamples = formatExamples("Cleanup examples:", assessment.riskExamples);
+    const headline = assessment.recommendation === "skip"
+        ? "Filtering may not be needed"
+        : assessment.recommendation === "light-filter"
+            ? "I recommend light cleanup filters"
+            : assessment.recommendation === "manual-review"
+                ? "Review filters or skip"
+                : "I recommend adding filters";
+    const agentIntent = assessment.recommendation === "skip"
+        ? `Codex is not recommending filters because the sample already looks clean. ${fitExamples} Add filters only for a narrow exclusion.`
+        : assessment.recommendation === "light-filter"
+            ? `Codex is recommending light filters because most rows fit, but ${cleanupCount} of ${assessment.sampleCount} look adjacent, unknown, or risky. ${riskExamples}`
+            : assessment.recommendation === "manual-review"
+                ? "Codex could not inspect the review rows, so it is pausing for a manual filter decision instead of assuming filters are required."
+                : `Codex is recommending filters because ${cleanupCount} of ${assessment.sampleCount} sampled rows look adjacent, unknown, or risky before message review. ${riskExamples}`;
+    const nextAction = assessment.recommendation === "skip"
+        ? "Skip filters or add a narrow cleanup rule"
+        : "Choose filters or skip";
     return {
         stage: "fit-message",
-        headline: "I recommend adding filters",
-        visibleState: `${reviewCopy} The browser is showing the filter-choice screen and sample rows.`,
-        agentIntent: "Codex is pausing on this sample because a mixed review list should be filtered before message review. Skip filters only if the visible rows already look clean.",
-        nextAction: "Choose filters or skip",
+        headline,
+        visibleState: `${reviewCopy} ${sampleCopy} The browser is showing the filter-choice screen.`,
+        agentIntent: agentIntent.trim(),
+        nextAction,
         safety: "This choice sets the filter path.",
         workerStatuses: {
             leadFitBuilder: "idle",
@@ -1011,7 +1139,7 @@ export const leadToolDefinitions = [
     },
     {
         name: "search_prospeo",
-        description: 'Search Prospeo for people using filters and optional domainFilterId. Requires get_provider_prompt({ provider: "prospeo" }) first. When targeting known accounts, call load_csv_domains for CSV-on-disk workflows or save_domain_filters for pasted/raw domain lists, then pass domainFilterId. Raw domain inputs and company-name targeting are NOT supported in this MCP tool. Strategy: start with 2-3 high-signal filters (title/seniority + industry or domainFilterId + headcount), then tighten one filter at a time. For security, AppSec, SOC, RevOps, Demand Gen, and similar function-specific lanes, do not widen with bare seniority labels like "Head" or "Director" alone; pair them with explicit function-title keywords and inspect the sample for off-function `Head of X` leakage. Prefer person location over company HQ unless HQ is explicitly needed. `campaignOfferId` routing rule: OMIT campaignOfferId ONLY in pre-mint Phase 84 `find leads` discovery mode (validating ICP before the commit gate). In every other context — post-mint lead additions, operator-driven searches on a live campaign, any search where the intent is to persist results to a specific campaign — you MUST pass campaignOfferId so the search shows up in that campaign\'s Contact Search panel. Post-mint create-campaign-v2 watch runs MUST pass currentStep: "prospeo". Omitting campaignOfferId post-mint orphans the search from the UI. Returns normalized results with pagination.',
+        description: 'Search Prospeo for people using filters and optional domainFilterId. Requires get_provider_prompt({ provider: "prospeo" }) first. Use Prospeo first for hiring-led targeting because it supports company_job_posting_hiring_for and company_job_posting_quantity; Sales Nav does not filter companies by hiring role. When targeting known accounts, call load_csv_domains for CSV-on-disk workflows or save_domain_filters for pasted/raw domain lists, then pass domainFilterId. Raw domain inputs and company-name targeting are NOT supported in this MCP tool. Strategy: start with 2-3 high-signal filters (title/seniority + industry or domainFilterId + headcount), add job-posting filters for hiring-led campaigns, then tighten one filter at a time. For security, AppSec, SOC, RevOps, Demand Gen, and similar function-specific lanes, do not widen with bare seniority labels like "Head" or "Director" alone; pair them with explicit function-title keywords and inspect the sample for off-function `Head of X` leakage. Prefer person location over company HQ unless HQ is explicitly needed. `campaignOfferId` routing rule: OMIT campaignOfferId ONLY in pre-mint Phase 84 `find leads` discovery mode (validating ICP before the commit gate). In every other context — post-mint lead additions, operator-driven searches on a live campaign, any search where the intent is to persist results to a specific campaign — you MUST pass campaignOfferId so the search shows up in that campaign\'s Contact Search panel. Post-mint create-campaign-v2 watch runs MUST pass currentStep: "prospeo". Omitting campaignOfferId post-mint orphans the search from the UI. Returns normalized results with pagination.',
         inputSchema: {
             type: "object",
             properties: {
@@ -1078,8 +1206,12 @@ export const leadToolDefinitions = [
                         company_job_posting_hiring_for: {
                             type: "array",
                             items: { type: "string" },
+                            description: "Company hiring filter. Use for hiring-led campaigns; values are open-role themes such as SDR, Account Executive, Growth, RevOps, Demand Generation, GTM Engineer, or AI Engineer.",
+                        },
+                        company_job_posting_quantity: {
+                            ...prospeoRangeFilterSchema,
+                            description: "Active job posting count range. Use min when the campaign needs companies with a real hiring push.",
                         },
-                        company_job_posting_quantity: prospeoRangeFilterSchema,
                         company_headcount_by_department: {
                             type: "array",
                             items: prospeoCompanyHeadcountByDepartmentSchema,
@@ -1192,7 +1324,7 @@ export const leadToolDefinitions = [
     },
     {
         name: "import_leads",
-        description: "Create/select a lead list and start the provider import job. Requires provider prompt preflight via get_provider_prompt for the active provider. On success, this tool owns moving the watched campaign to confirm-lead-list with import/progress narration after a leadListId/jobId exists; do not call update_campaign to fix that step. After the user confirms the list, call confirm_lead_list.",
+        description: "Create/select a source lead list and start the provider import/export job. Requires provider prompt preflight via get_provider_prompt for the active provider. For Sales Nav/Prospeo, targetLeadCount is the approved source-list export count computed from projected good-fit math, not the later bounded review-batch size. On success, this tool owns moving the watched campaign to confirm-lead-list with import/progress narration after a leadListId/jobId exists; do not call update_campaign to fix that step. After the user confirms the source list, call confirm_lead_list to clone only the bounded review batch into the campaign table.",
         inputSchema: {
             type: "object",
             properties: {
@@ -1219,7 +1351,7 @@ export const leadToolDefinitions = [
                 },
                 targetLeadCount: {
                     type: "number",
-                    description: "Target number of leads (max per provider from config)",
+                    description: "Provider source-list target count (max per provider from config). For Sales Nav/Prospeo this should be the export/materialization count needed to hit projected good-fit goals, not the 25-row review-batch size.",
                 },
                 mode: {
                     type: "string",
@@ -2741,6 +2873,20 @@ export async function confirmLeadList(input) {
             await Promise.all(batch.map((rowId) => api.delete(`/api/v3/workflow-tables/${campaignTableId}/rows/${rowId}`)));
         }
     }
+    let reviewSampleRows = [];
+    if (campaignTableId && effectiveCurrentStep === "filter-choice") {
+        try {
+            const sampleLimit = Math.min(Math.max(keptReviewRowCount, 1), 25);
+            const sample = await getTableRowsMinimal(campaignTableId, {
+                limit: sampleLimit,
+                page: 1,
+            });
+            reviewSampleRows = Array.isArray(sample?.rows) ? sample.rows : [];
+        }
+        catch {
+            reviewSampleRows = [];
+        }
+    }
     // Persist currentStep if the caller asked for it. Do NOT touch
     // selectedLeadListId here: the campaign table id is already saved to
     // CampaignOffer.workflowTableId by /api/v3/campaign-builder/import-leads,
@@ -2758,6 +2904,7 @@ export async function confirmLeadList(input) {
                     watchNarration: buildFilterChoiceWatchNarration({
                         sourceLeadCount: leadListRowCount,
                         reviewRowCount: keptReviewRowCount,
+                        sampleRows: reviewSampleRows,
                     }),
                 }
                 : {}),
@@ -2800,8 +2947,8 @@ export async function confirmLeadList(input) {
             : undefined,
         message: requestedTargetLeadCount !== null &&
             leadListRowCount > requestedTargetLeadCount
-            ? `I found ${leadListRowCount} source candidates and imported the first ${Math.min(importedRowCount, requestedTargetLeadCount)} into the campaign review table. The watched campaign is now on filter-choice with sample assessment copy; announce Message Draft Builder background start only if that branch has actually started.`
-            : "Lead list imported into the campaign review table. The watched campaign is now on filter-choice with sample assessment copy; announce Message Draft Builder background start only if that branch has actually started.",
+            ? `I found ${leadListRowCount} source candidates and imported the first ${Math.min(importedRowCount, requestedTargetLeadCount)} into the campaign review table. The watched campaign is now on filter-choice; ask add filters vs skip filters before loading post-lead workers.`
+            : "Lead list imported into the campaign review table. The watched campaign is now on filter-choice; ask add filters vs skip filters before loading post-lead workers.",
     };
 }
 export function getProviderPrompt(input) {

package/dist/tools/prompts.d.ts

@@ -134,6 +134,7 @@ export interface PostFindLeadsScoutRegistryResponse {
         forbiddenStarts: string[];
         runtimeProofTransport: string;
         runtimeProofRequiredFields: string[];
+        promptRequired?: string;
         basisFields: string[];
         compactOutputFields: string[];
         reusePolicy?: string;

package/dist/tools/prompts.js

@@ -315,6 +315,7 @@ export function getPostFindLeadsScoutRegistry() {
                 "basis.workflowTableId",
                 "basis.reviewBatchRowHash or basis.reviewBatchRowIds",
             ],
+            promptRequired: 'get_subskill_prompt({ subskillName: "generate-messages", offset, limit }) until hasMore=false',
             basisFields: [
                 "campaign revision or updatedAt",
                 "brief hash",
@@ -339,9 +340,9 @@ export function getPostFindLeadsScoutRegistry() {
             reusePolicy: "The first completed Message Draft Builder recommendation remains the default review candidate. Later Lead Fit Builder, Filter Leads, enrichment, or rubric completion may make an enriched rewrite available, but does not automatically retry or replace the initial draft unless campaign/brief/source/list/table/review-batch identity mismatches or the initial output failed.",
         },
         usage: {
-            codex: "After confirm_lead_list imports a non-empty bounded review batch and get_rows_minimal proves rows for workflowTableId, spawn the returned Message Draft Builder scout name immediately when the current Codex host exposes that custom agent. Spawn the Lead Fit Builder scout only after the user chooses filters.",
-            claude: "After confirm_lead_list imports a non-empty bounded review batch and get_rows_minimal proves rows for workflowTableId, invoke the returned Message Draft Builder Task/Agent immediately when the current Claude session lists that agent. Invoke the Lead Fit Builder only after the user chooses filters.",
-            parentThreadRule: "Named agents are optional acceleration. If they are absent, do not customer-surface install status; the main thread still orchestrates filter and message branches from CampaignOffer state, selected source state, workflowTableId, and imported review-batch rows. Local markdown/json files are not normal-path inputs. The message drafting branch starts at the review-batch import gate, must load the full generate-messages prompt, must read live campaign/review-batch state through scoped MCP/product tools, and must reject mismatched selectedLeadListId/workflowTableId/campaign/workspace input. Template approval in the parent thread uses the create-campaign-v2 message-review safety gate; do not load the full generate-messages prompt for approval-only review. Join before message review. Do not automatically rerun Message Draft Builder after filters/enrichment finish; show the initial draft by default and offer an enriched rewrite only with explicit user opt-in.",
+            codex: "After confirm_lead_list imports a non-empty bounded review batch and get_rows_minimal proves rows for workflowTableId, ask the filter-choice question immediately. Do not spawn returned post-lead scout names before that question. Once the user answers, spawn Message Draft Builder from the same campaign/table basis. If the user chooses filters, also spawn Lead Fit Builder, move to Filter Rules, save rubrics, then keep the browser on Filter Leads while the message recommendation is reviewed. If filters are skipped, move to Messages/message review.",
+            claude: "After confirm_lead_list imports a non-empty bounded review batch and get_rows_minimal proves rows for workflowTableId, ask the filter-choice question immediately. Do not invoke returned post-lead Task/Agent names before that question. Once the user answers, invoke Message Draft Builder from the same campaign/table basis. If the user chooses filters, also invoke Lead Fit Builder, move to Filter Rules, save rubrics, then keep the browser on Filter Leads while the message recommendation is reviewed. If filters are skipped, move to Messages/message review.",
+            parentThreadRule: 'Named agents are optional acceleration, but message drafting is not optional. If post-find-leads-message-scout is available, run it as the background Message Draft Builder after the filter-choice answer. If it is absent, do not customer-surface install status; the main thread must execute the same message branch from CampaignOffer state, selected source state, workflowTableId, and imported review-batch rows. Local markdown/json files are not normal-path inputs. The filter-choice question is the first post-import user gate; do not load post-lead registries, filter references, or the full generate-messages prompt before it. Message drafting starts after the filter-choice answer, must load get_subskill_prompt({ subskillName: "generate-messages", offset, limit }) until hasMore=false, must read live campaign/review-batch state through scoped MCP/product tools, and must reject mismatched selectedLeadListId/workflowTableId/campaign/workspace input. On the filter path, keep the browser on Filter Rules then Filter Leads while the background message agent prepares the template; enrichment/filtering/Generate Message cells wait for template approval. On the skip path, move to Messages/message review and wait for template approval before enrichment. Do not render message review from checklist or shortcut instructions; message review requires a messageDraftRecommendation whose basis proves the full generate-messages prompt ran for the current campaign/table/review batch. Do not automatically rerun Message Draft Builder after filters/enrichment finish; show the initial draft by default and offer an enriched rewrite only with explicit user opt-in.',
         },
     };
 }

package/dist/tools/registry.d.ts

@@ -2557,10 +2557,11 @@ export declare const allTools: ({
                         items: {
                             type: string;
                         };
+                        description: string;
                     };
                     company_job_posting_quantity: {
-                        type: string;
                         description: string;
+                        type: string;
                         properties: {
                             min: {
                                 type: string;
@@ -3702,6 +3703,7 @@ export declare const allTools: ({
             tableId?: undefined;
             targetCount?: undefined;
             minPassedCount?: undefined;
+            minMessagesCount?: undefined;
             timeoutMs?: undefined;
             intervalMs?: undefined;
             includeRows?: undefined;
@@ -3735,6 +3737,7 @@ export declare const allTools: ({
             tableId?: undefined;
             targetCount?: undefined;
             minPassedCount?: undefined;
+            minMessagesCount?: undefined;
             timeoutMs?: undefined;
             intervalMs?: undefined;
             includeRows?: undefined;
@@ -3793,6 +3796,7 @@ export declare const allTools: ({
             tableId?: undefined;
             targetCount?: undefined;
             minPassedCount?: undefined;
+            minMessagesCount?: undefined;
             timeoutMs?: undefined;
             intervalMs?: undefined;
             includeRows?: undefined;
@@ -3851,6 +3855,7 @@ export declare const allTools: ({
             tableId?: undefined;
             targetCount?: undefined;
             minPassedCount?: undefined;
+            minMessagesCount?: undefined;
             timeoutMs?: undefined;
             intervalMs?: undefined;
             includeRows?: undefined;
@@ -3911,6 +3916,7 @@ export declare const allTools: ({
             tableId?: undefined;
             targetCount?: undefined;
             minPassedCount?: undefined;
+            minMessagesCount?: undefined;
             timeoutMs?: undefined;
             intervalMs?: undefined;
             includeRows?: undefined;
@@ -3941,6 +3947,7 @@ export declare const allTools: ({
             tableId?: undefined;
             targetCount?: undefined;
             minPassedCount?: undefined;
+            minMessagesCount?: undefined;
             timeoutMs?: undefined;
             intervalMs?: undefined;
             includeRows?: undefined;
@@ -3971,6 +3978,7 @@ export declare const allTools: ({
             tableId?: undefined;
             targetCount?: undefined;
             minPassedCount?: undefined;
+            minMessagesCount?: undefined;
             timeoutMs?: undefined;
             intervalMs?: undefined;
             includeRows?: undefined;

package/dist/tools/rows.d.ts

@@ -120,6 +120,7 @@ export declare function getTableRows(tableId: string, options?: {
 export declare function getTableRowsMinimal(tableId: string, options?: {
     limit?: number;
     page?: number;
+    includeMessages?: boolean;
 }): Promise<RowsResponse>;
 export declare function getRows(tableId: string, rowIds: string | string[]): Promise<{
     rows: FullRow[];

package/dist/tools/rows.js

@@ -90,6 +90,8 @@ export async function getTableRowsMinimal(tableId, options) {
     if (options?.page)
         params.set("page", String(options.page));
     params.set("minimal", "1");
+    if (options?.includeMessages)
+        params.set("includeMessages", "1");
     const queryString = params.toString();
     const path = `/api/v3/mcp/tables/${tableId}/rows${queryString ? `?${queryString}` : ""}`;
     return api.get(path);