@sellable/mcp 0.1.134 → 0.1.136

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

@@ -11,11 +11,14 @@ campaign state. The main thread owns approval and campaign writes.
 Use the live campaign inputs supplied by the parent thread:
 
 - `campaignId`
+- campaign revision or `campaignUpdatedAt`
 - `campaignBrief` / campaign brief content model
 - selected source decision and provider state
 - `selectedLeadListId` or selected source list context
 - `workflowTableId`
-- imported review-batch rows from that selected list
+- imported review-batch rows from that selected list, including row IDs and a
+  review-batch row hash when available
+- filter basis at branch start: `pending`, `use-filters`, or `skip-filters`
 - any already-saved fit/rubric result summaries supplied by the parent
 
 Do not require or hunt for `brief.md`, `lead-review.md`, or `lead-sample.json`.
@@ -23,6 +26,13 @@ Those files are optional debug context only when the parent explicitly provides
 them. Never inspect the product database directly, never run `psql`, and never
 read stale local markdown files to reconstruct campaign state.
 
+All live reads must come from scoped MCP/product tools by campaign and
+workspace, such as `get_campaign`, `get_campaign_context`, and
+`get_rows_minimal({ tableId: workflowTableId })`, or from equivalent parent
+thread payloads. Reject the task as `blocked` if the campaign id, workspace,
+`selectedLeadListId`, `workflowTableId`, or review-batch row ids do not match
+the branch input.
+
 ## Required First Steps
 
 1. Load the full long-form generate-messages prompt:
@@ -45,11 +55,31 @@ Return the following to the parent thread:
 - one rendered good-fill sample for a plausible passing review-batch row
 - one omit/fallback sample when the row signal is not safe
 - pass/fail notes against the generate-messages quality gates
+- compact runtime status: `ready`, `blocked`, `retry-needed`, or `stale`
+- basis token containing campaign revision/updatedAt, brief hash,
+  `selectedLeadListId`, `workflowTableId`, review-batch row ids/hash, filter
+  choice, and rubric/filter basis when present
+- output timestamp/hash and any retry/error detail
 
 Write `message-validation.md`, `message-prep.md`, or
 `message-candidate-drafts.md` only when the parent explicitly asks for debug
 artifacts. Normal live campaign runs can return the same content directly.
 
+When reporting branch runtime proof, use this shape under
+`watchNarration.workerDetails.messageDraftBuilder`:
+
+- `statusSource`: `branch` or `parent-thread-fallback`
+- `status`: `branch-running`, `fallback-active`, `spawn-failed`,
+  `fallback-superseded`, `branch-superseded`, `ready`, `blocked`,
+  `retry-needed`, or `stale`
+- `runId` or `fallbackId`
+- `startedAt` and `updatedAt`
+- `basisToken` and `basis`
+- optional `compactOutputRef`, `compactOutput`, and `error`
+
+Do not tell the UI to show Message Draft Builder as running unless this proof
+exists and points at the current non-empty bounded review batch.
+
 ## Hard Rules
 
 - Do not call product Generate Message cells. This worker drafts the template

package/agents/registry.json

@@ -225,10 +225,10 @@
       ],
       "ownership": "proof inventory, token strategy, angle drafting, skeptical-prospect review, and selected winner only",
       "codex": {
-        "description": "Message Draft Builder for campaign-backed template proposals and sample messages after source approval.",
+        "description": "Message Draft Builder for campaign-backed template proposals after confirm_lead_list imports a non-empty bounded review batch.",
         "model": "gpt-5.5",
         "modelReasoningEffort": "high",
-        "sandboxMode": "workspace-write",
+        "sandboxMode": "read-only",
         "nicknameCandidates": [
           "Message Draft Builder",
           "Draft Builder",
@@ -236,19 +236,17 @@
         ]
       },
       "claude": {
-        "description": "Use proactively as Message Draft Builder after lead source approval to propose an approved-template candidate and sample message from campaign state.",
+        "description": "Use proactively as Message Draft Builder after confirm_lead_list imports a non-empty bounded review batch; load the full generate-messages prompt and draft only from scoped campaign/tool state.",
         "model": "inherit",
         "background": true,
         "maxTurns": 10,
         "color": "magenta",
         "tools": [
           "Read",
-          "Write",
-          "Edit",
-          "Grep",
-          "Glob",
           "mcp__sellable__get_subskill_prompt",
-          "mcp__sellable__get_subskill_asset"
+          "mcp__sellable__get_campaign",
+          "mcp__sellable__get_campaign_context",
+          "mcp__sellable__get_rows_minimal"
         ]
       }
     }

package/agents/source-scout-linkedin-engagement.md

@@ -8,9 +8,10 @@ Required first step:
   draft `campaignOfferId`, call `get_provider_prompt({ provider:
 "signal-discovery", campaignOfferId, confirmed: true })` and include that same
   `campaignOfferId` plus `currentStep: "signal-discovery"` in `search_signals`
-  so the user can watch source work in the campaign UI. Treat that as a
-  campaign-attached persisted search; do not run a post-mint search without the
-  campaign ID. If no campaign ID is supplied, run campaignless preview mode.
+  so the owning search route can show the source lane with current find-leads
+  narration and user options. Treat that as a campaign-attached persisted
+  search; do not run a post-mint search without the campaign ID. If no campaign
+  ID is supplied, run campaignless preview mode.
 
 Use the inherited Sellable MCP tools when available:
 
@@ -44,6 +45,8 @@ sales Claude`; do not treat broad anchor-only lanes like `Claude Code`, `MCP`,
 campaignOfferId, selectionMode: "replace", selections, headlineICPCriteria,
 currentStep: "signal-discovery" })` before sampling so the watched Signal
    Discovery table shows the promoted posts and the exact posts being tested.
+   Do not move the campaign to `confirm-lead-list`; `import_leads` owns that
+   visible transition after source approval.
 6. Fetch or sample engagers for promoted posts and score rough ICP fit from
    visible headline/display-name cues only. Do not enrich people during
    viability estimation.

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

@@ -8,10 +8,12 @@ Required first step:
   draft `campaignOfferId`, call `get_provider_prompt({ provider: "prospeo",
 campaignOfferId, confirmed: true })` and include that same `campaignOfferId`
   plus `currentStep: "prospeo"` in `search_prospeo` so the user can watch source
-  work in the campaign UI. If no campaign ID is supplied, run campaignless
-  preview mode. Treat post-mint
+  work in the campaign UI with source-lane narration owned by the search route.
+  If no campaign ID is supplied, run campaignless preview mode. Treat post-mint
   searches with `campaignOfferId` as campaign-attached persisted search tabs;
   do not run a live campaign search without the campaign ID.
+  Do not move the campaign to `confirm-lead-list`; `import_leads` owns that
+  visible transition after source approval.
 
 Use the inherited Sellable MCP tools when available:
 

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

@@ -8,10 +8,12 @@ Required first step:
   draft `campaignOfferId`, call `get_provider_prompt({ provider: "sales-nav",
 campaignOfferId, confirmed: true })` and include that same `campaignOfferId` in
   `search_sales_nav` with `currentStep: "sales-nav"` so the user can watch
-  source work in the campaign UI. If no campaign ID is supplied, run
-  campaignless preview mode. Treat post-mint
+  source work in the campaign UI with source-lane narration owned by the search
+  route. If no campaign ID is supplied, run campaignless preview mode. Treat post-mint
   searches with `campaignOfferId` as campaign-attached persisted search tabs;
   do not run a live campaign search without the campaign ID.
+  Do not move the campaign to `confirm-lead-list`; `import_leads` owns that
+  visible transition after source approval.
 
 Use the inherited Sellable MCP tools when available:
 

package/dist/tools/campaigns.d.ts

@@ -568,6 +568,12 @@ export interface CreateCampaignResult {
     watchUrl: string;
     resumed?: boolean;
     warning?: string;
+    approvalGate: {
+        kind: "brief-review";
+        mustShowBeforeQuestion: string[];
+        questionToolBlockedUntilShown: true;
+    };
+    nextStep: string;
 }
 export declare function createCampaign(input: CreateCampaignInput): Promise<CreateCampaignResult>;
 export interface UpdateCampaignResult {

package/dist/tools/campaigns.js

@@ -658,6 +658,20 @@ function deriveOfferValue(name, briefContent) {
         return heading.slice(0, 120);
     return lines[0].slice(0, 120);
 }
+function buildBriefApprovalGate(watchUrl) {
+    return {
+        approvalGate: {
+            kind: "brief-review",
+            mustShowBeforeQuestion: [
+                "the full campaignBrief markdown the user is approving",
+                `Watch link: ${watchUrl}`,
+            ],
+            questionToolBlockedUntilShown: true,
+        },
+        nextStep: "Before asking for approval, render the full campaign brief in normal chat text and then print " +
+            `Watch link: ${watchUrl}. Do not call AskUserQuestion or request_user_input until the user has seen both the brief content and this watch link.`,
+    };
+}
 export async function createCampaign(input) {
     const api = getApi();
     const config = getConfig();
@@ -675,6 +689,7 @@ export async function createCampaign(input) {
             createdAt: existing.createdAt,
             currentStep: existing.currentStep ?? null,
             watchUrl,
+            ...buildBriefApprovalGate(watchUrl),
             resumed: true,
             warning: hasCreateFields
                 ? "campaignId provided; resume mode ignores create fields."
@@ -790,6 +805,7 @@ export async function createCampaign(input) {
         createdAt: result.createdAt,
         currentStep: result.currentStep,
         watchUrl,
+        ...buildBriefApprovalGate(watchUrl),
     };
 }
 export async function updateCampaign(campaignId, input) {
@@ -827,17 +843,6 @@ export async function updateCampaign(campaignId, input) {
             }
         }
     }
-    // Hotfix: when rubrics are saved with enableICPFilters, ensure the UI step
-    // is set to "filter-rules" so the frontend can render the filter state.
-    // Without this, the MCP caller can save rubrics + enableICPFilters on the
-    // backend while the UI currentStep points somewhere else (e.g. "messages"),
-    // causing the UI to be out of sync with the backend filter state.
-    if (updates.enableICPFilters === true &&
-        Array.isArray(updates.rubric) &&
-        updates.rubric.length > 0 &&
-        !updates.currentStep) {
-        updates.currentStep = "filter-rules";
-    }
     let result = null;
     let senderRouteResult = null;
     const shouldPreserveInteractionModePut = typeof interactionMode === "string" &&

package/dist/tools/leads.d.ts

@@ -2540,6 +2540,20 @@ export declare function confirmLeadList(input: ConfirmLeadListInput): Promise<{
         remainingRowCount?: number | null;
         rowCount?: number | null;
     };
+    messageDraftBuilder: {
+        firstAllowedStartPoint: string;
+        startAllowed: boolean;
+        requiredBeforeRunningCopy: string;
+        requiredLiveInputs: {
+            campaignOfferId: string;
+            selectedLeadListId: string;
+            workflowTableId: string | null;
+            reviewBatchRowIds: string[];
+            reviewBatchRowCount: number;
+        };
+        branchBasisFields: string[];
+        promptRequired: string;
+    };
     boundedReviewBatch: {
         requestedTargetLeadCount: number;
         importedRowCount: number;

package/dist/tools/leads.js

@@ -429,6 +429,81 @@ function summarizeSignalSearchResponse(response) {
         ],
     };
 }
+function buildSourceImportWatchNarration({ provider, selectedPostCount, estimatedEngagers, targetLeadCount, }) {
+    const providerLabel = provider === "signal-discovery"
+        ? "Signal Leads"
+        : provider === "sales-nav"
+            ? "Sales Nav"
+            : provider === "prospeo"
+                ? "Prospeo"
+                : "Apollo";
+    const sourceDetail = provider === "signal-discovery"
+        ? `${selectedPostCount ?? "selected"} approved post${selectedPostCount === 1 ? "" : "s"}${typeof estimatedEngagers === "number"
+            ? ` with about ${estimatedEngagers.toLocaleString("en-US")} engagers`
+            : ""}`
+        : `the approved ${providerLabel} source`;
+    const targetDetail = typeof targetLeadCount === "number"
+        ? ` Targeting ${targetLeadCount.toLocaleString("en-US")} source leads before the bounded review batch is cloned.`
+        : "";
+    return {
+        stage: "review-batch",
+        headline: provider === "signal-discovery"
+            ? "Scraping source leads from posts"
+            : "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: "No enrichment, sequence, or sending starts during source import.",
+    };
+}
+function buildSourceImportRecoveryWatchNarration(args) {
+    const providerLabel = args.provider === "signal-discovery"
+        ? "Signal Leads"
+        : args.provider === "sales-nav"
+            ? "Sales Nav"
+            : args.provider === "prospeo"
+                ? "Prospeo"
+                : "source";
+    const reasonCopy = args.reason === "failed"
+        ? `${providerLabel} import failed before review rows were ready.`
+        : args.reason === "zero"
+            ? `${providerLabel} import finished without usable review rows.`
+            : args.reason === "timeout"
+                ? `${providerLabel} import has not produced review rows within the wait window.`
+                : `${providerLabel} import is still materializing source rows.`;
+    return {
+        stage: "review-batch",
+        headline: args.reason === "failed" || args.reason === "zero"
+            ? "Source import needs attention"
+            : "Source import still running",
+        visibleState: `${reasonCopy} The browser should stay on the lead import/review-batch screen.`,
+        agentIntent: "Codex is holding the campaign before filter-choice until a non-empty bounded review batch exists.",
+        nextAction: args.reason === "failed" || args.reason === "zero"
+            ? "Retry the import or change the approved source"
+            : "Wait again, retry readiness, or change the source",
+        safety: "No enrichment, filtering, message cells, sequence, or sending has started.",
+    };
+}
+function buildFilterChoiceWatchNarration({ sourceLeadCount, reviewRowCount, }) {
+    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}.`;
+    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",
+        safety: "No enrichment, filtering, Generate Message cells, sequence, or sending starts from this choice.",
+        workerStatuses: {
+            leadFitBuilder: "idle",
+            messageDraftBuilder: "idle",
+        },
+    };
+}
 function normalizeImportProvider(provider) {
     if (provider === "apollo-ai" || provider === "apollo")
         return "apollo";
@@ -960,7 +1035,7 @@ 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. Sets selectedLeadListId to the lead list so the user can review it. Does NOT import into the campaign table. After the user confirms the list, call confirm_lead_list.",
+        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.",
         inputSchema: {
             type: "object",
             properties: {
@@ -1053,7 +1128,7 @@ export const leadToolDefinitions = [
     },
     {
         name: "confirm_lead_list",
-        description: "After the user confirms the lead list looks good, import it into the campaign table (clone). selectedLeadListId remains the source lead list; workflowTableId is the campaign table. Recommended post-confirm order: update_campaign(currentStep='filter-choice') -> wait_for_campaign_table_ready -> get_rows_minimal.",
+        description: "After the user confirms the lead list looks good, import the bounded review batch into the campaign table. This tool owns moving the watched campaign to filter-choice with sample-assessment narration only after non-empty review rows exist; do not call update_campaign to fix filter-choice afterward. selectedLeadListId remains the source lead list; workflowTableId is the campaign table.",
         inputSchema: {
             type: "object",
             properties: {
@@ -1067,7 +1142,7 @@ export const leadToolDefinitions = [
                 },
                 currentStep: {
                     type: ["string", "null"],
-                    description: 'Headless workflow step ID. If omitted, defaults to "confirm-lead-list". Pass null to skip auto-advance.',
+                    description: 'Headless workflow step ID. If omitted, defaults to "filter-choice" after review rows exist. Pass null to skip auto-advance.',
                 },
                 campaignName: {
                     type: "string",
@@ -2061,6 +2136,17 @@ export async function importLeads(input) {
         await api.put(`/api/v2/campaign-offers/${campaignOfferId}`, {
             selectedLeadListId: result.tableId,
             ...(shouldSetCurrentStep ? { currentStep: effectiveCurrentStep } : {}),
+            ...(shouldSetCurrentStep
+                ? {
+                    watchNarration: buildSourceImportWatchNarration({
+                        provider: "signal-discovery",
+                        selectedPostCount: postsToScrape.length,
+                        estimatedEngagers: result.estimatedEngagers,
+                        targetLeadCount: normalizePositiveInteger(targetEngagerCount) ??
+                            result.estimatedEngagers,
+                    }),
+                }
+                : {}),
         });
         return {
             provider: "signal-discovery",
@@ -2075,7 +2161,7 @@ export async function importLeads(input) {
             targetLeadCount: cappedTargetLeadCount ?? null,
             message: `Started scraping ${postsToScrape.length} posts (~${result.estimatedEngagers} engagers). Leads will appear as scraping completes.${importSelection.limited
                 ? ` Limited from ${uniqueSelectedPosts.length} selected posts by the approved source-capacity scrape plan.`
-                : ""}`,
+                : ""} The watched campaign has been moved to confirm-lead-list with import progress copy; do not call update_campaign to fix that step.`,
         };
     }
     // === SALES NAV / PROSPEO FLOW ===
@@ -2172,6 +2258,14 @@ export async function importLeads(input) {
     await api.put(`/api/v2/campaign-offers/${campaignOfferId}`, {
         selectedLeadListId: leadListId,
         ...(shouldSetCurrentStep ? { currentStep: effectiveCurrentStep } : {}),
+        ...(shouldSetCurrentStep
+            ? {
+                watchNarration: buildSourceImportWatchNarration({
+                    provider,
+                    targetLeadCount: cappedTargetLeadCount ?? null,
+                }),
+            }
+            : {}),
     });
     return {
         provider,
@@ -2180,7 +2274,7 @@ export async function importLeads(input) {
         jobResult,
         jobId,
         targetLeadCount: cappedTargetLeadCount ?? null,
-        message: "Import started. Review the lead list as it fills; once it looks good, confirm to import into the campaign table.",
+        message: "Import started and the watched campaign has been moved to confirm-lead-list with import progress copy. Review the lead list as it fills; once it looks good, call confirm_lead_list. Do not call update_campaign to fix the import step.",
     };
 }
 export async function cancelLeadImport(input) {
@@ -2332,12 +2426,24 @@ export async function confirmLeadList(input) {
     }
     if (!readiness.ready) {
         if (readiness.reason === "missing_job_id") {
-            throw new Error("Import job ID is missing. Keep the campaign at confirm-lead-list; provide the jobId, retry readiness, cancel the import, or re-run-source before confirming.");
+            const recoveryNarration = buildSourceImportRecoveryWatchNarration({
+                reason: "pending",
+                provider: resolvedProvider,
+            });
+            throw new Error(`${recoveryNarration.headline}: Import job ID is missing. Keep the campaign at confirm-lead-list; provide the jobId, retry readiness, cancel the import, or re-run-source before confirming. ${recoveryNarration.safety}`);
         }
         if (readiness.reason === "import_failed") {
-            throw new Error("Import failed. Keep the campaign at confirm-lead-list; retry the provider import, cancel it, or re-run-source before confirming.");
+            const recoveryNarration = buildSourceImportRecoveryWatchNarration({
+                reason: "failed",
+                provider: resolvedProvider,
+            });
+            throw new Error(`${recoveryNarration.headline}: Import failed. Keep the campaign at confirm-lead-list; retry the provider import, cancel it, or re-run-source before confirming. ${recoveryNarration.safety}`);
         }
-        throw new Error("Import still in progress. Keep the campaign at confirm-lead-list; retry readiness, cancel the import, or re-run-source before launching post-import scouts.");
+        const recoveryNarration = buildSourceImportRecoveryWatchNarration({
+            reason: "timeout",
+            provider: resolvedProvider,
+        });
+        throw new Error(`${recoveryNarration.headline}: Import still in progress. Keep the campaign at confirm-lead-list; retry readiness, cancel the import, or re-run-source before launching post-import scouts. ${recoveryNarration.safety}`);
     }
     const isTerminalAccessError = (error) => error instanceof SellableApiError && [401, 403, 404].includes(error.status);
     const formatTerminalAccessError = (error) => {
@@ -2353,7 +2459,7 @@ export async function confirmLeadList(input) {
         campaignName,
         keepInSync,
         ...(typeof targetLeadCount === "number" ? { targetLeadCount } : {}),
-        ...(shouldSetCurrentStep ? { currentStep: effectiveCurrentStep } : {}),
+        currentStep: null,
     })
         .catch((error) => {
         if (isTerminalAccessError(error)) {
@@ -2375,9 +2481,15 @@ export async function confirmLeadList(input) {
         ? importedRowIds.length
         : typeof importResult.rowCount === "number"
             ? importResult.rowCount
-            : typeof importResult.leadsImported === "number"
-                ? importResult.leadsImported
-                : 0;
+            : typeof importResult.leadsImported === "number" &&
+                typeof importResult.leadsSkipped === "number"
+                ? importResult.leadsImported + importResult.leadsSkipped
+                : typeof importResult.leadsImported === "number"
+                    ? importResult.leadsImported
+                    : 0;
+    const keptReviewRowCount = requestedTargetLeadCount !== null
+        ? Math.min(importedRowCount, requestedTargetLeadCount)
+        : importedRowCount;
     const remainingRowCount = typeof importResult.remainingRowCount === "number"
         ? importResult.remainingRowCount
         : 0;
@@ -2392,7 +2504,11 @@ export async function confirmLeadList(input) {
         throw new Error("Campaign review rows are still importing. Stay on lead import until the bounded campaign table is ready, then retry confirm_lead_list or wait_for_campaign_table_ready.");
     }
     if (importedRowCount <= 0) {
-        throw new Error("No usable review rows were kept for the campaign table. Tighten or change the source before continuing.");
+        const recoveryNarration = buildSourceImportRecoveryWatchNarration({
+            reason: "zero",
+            provider: resolvedProvider,
+        });
+        throw new Error(`${recoveryNarration.headline}: No usable review rows were kept for the campaign table. Retry the import or change the approved source before continuing. ${recoveryNarration.safety}`);
     }
     if (campaignTableId && overflowRowIds.length > 0) {
         const deleteBatchSize = 25;
@@ -2413,12 +2529,41 @@ export async function confirmLeadList(input) {
     if (shouldSetCurrentStep) {
         await api.put(`/api/v2/campaign-offers/${campaignOfferId}`, {
             currentStep: effectiveCurrentStep,
+            ...(effectiveCurrentStep === "filter-choice"
+                ? {
+                    watchNarration: buildFilterChoiceWatchNarration({
+                        sourceLeadCount: leadListRowCount,
+                        reviewRowCount: keptReviewRowCount,
+                    }),
+                }
+                : {}),
         });
     }
     return {
         sourceLeadListId: resolvedLeadListId,
         campaignTableId: campaignTableId ?? null,
         importResult,
+        messageDraftBuilder: {
+            firstAllowedStartPoint: "confirm_lead_list",
+            startAllowed: true,
+            requiredBeforeRunningCopy: "Persist watchNarration.workerDetails.messageDraftBuilder with runId or fallbackId, statusSource, status, startedAt, updatedAt, and current basis before showing Message Draft Builder as In Progress.",
+            requiredLiveInputs: {
+                campaignOfferId,
+                selectedLeadListId: resolvedLeadListId,
+                workflowTableId: campaignTableId ?? null,
+                reviewBatchRowIds: importedRowIds.slice(0, keptReviewRowCount),
+                reviewBatchRowCount: keptReviewRowCount,
+            },
+            branchBasisFields: [
+                "campaign revision or updatedAt",
+                "brief hash",
+                "selectedLeadListId",
+                "workflowTableId",
+                "bounded review-batch row ids/hash",
+                "filter choice at branch start",
+            ],
+            promptRequired: 'Load get_subskill_prompt({ subskillName: "generate-messages", offset, limit }) until hasMore=false before drafting.',
+        },
         boundedReviewBatch: requestedTargetLeadCount !== null
             ? {
                 requestedTargetLeadCount,
@@ -2431,8 +2576,8 @@ 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.`
-            : "Lead list imported into the campaign review table. Next: sample rows with get_rows_minimal or continue into filter work.",
+            ? `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.",
     };
 }
 export function getProviderPrompt(input) {

package/dist/tools/processing.js

@@ -17,7 +17,7 @@ export const processingToolDefinitions = [
                 },
                 currentStep: {
                     type: ["string", "null"],
-                    description: "Headless workflow step ID",
+                    description: "Headless workflow step ID only when paired with the visible watch narration expected by the campaign-offers API; do not use as a separate step-only fixup after rubric, sender, sequence, or start work.",
                 },
                 enableICPFilters: {
                     type: "boolean",

package/dist/tools/prompts.d.ts

@@ -127,10 +127,20 @@ export interface PostFindLeadsScoutRegistryResponse {
         afterAllComplete: boolean;
         requiredArtifacts?: string[];
         requiredState?: string[];
+        noFilterRequiredState?: string[];
+        skipFilterRule?: string;
         optionalDebugArtifacts?: string[];
         show: string[];
         nextStage: string;
     };
+    messageDraftBranchContract?: {
+        firstAllowedStart: string;
+        forbiddenStarts: string[];
+        runtimeProofTransport: string;
+        runtimeProofRequiredFields: string[];
+        basisFields: string[];
+        compactOutputFields: string[];
+    };
     usage: {
         codex: string;
         claude: string;

package/dist/tools/prompts.js

@@ -256,10 +256,12 @@ export function getPostFindLeadsScoutRegistry() {
         trigger: "find_leads_source_approved",
         requiredInputs: [
             "campaignId",
+            "campaign revision or campaignUpdatedAt",
             "campaignBrief",
             "source decision and selectedLeadList/source state",
             "workflowTableId",
-            "imported review-batch rows from selectedLeadList",
+            "imported review-batch rows from selectedLeadList with row ids/hash",
+            "filter choice and filter/rubric basis when present",
         ],
         optionalDebugArtifacts: ["brief.md", "lead-review.md", "lead-sample.json"],
         scouts: scouts.map((agent) => ({
@@ -290,14 +292,56 @@ export function getPostFindLeadsScoutRegistry() {
         joinGate: {
             afterAllComplete: true,
             requiredState: ["leadScoringRubrics", "messageDraftRecommendation"],
+            noFilterRequiredState: ["messageDraftRecommendation"],
+            skipFilterRule: "leadScoringRubrics may be absent only when the user explicitly skips filters with enableICPFilters=false; messageDraftRecommendation is still required before template review.",
             optionalDebugArtifacts: ["lead-filter.md", "message-validation.md"],
             show: ["readable_filters_with_reasons", "sample_message_for_approval"],
             nextStage: "message-review",
         },
+        messageDraftBranchContract: {
+            firstAllowedStart: "after confirm_lead_list imports a non-empty bounded review batch and get_rows_minimal confirms rows for workflowTableId",
+            forbiddenStarts: [
+                "source recommendation",
+                "provider import job alone",
+                "zero-row review batch",
+            ],
+            runtimeProofTransport: "CampaignOffer.watchNarration.workerDetails.messageDraftBuilder",
+            runtimeProofRequiredFields: [
+                "runId or fallbackId",
+                "statusSource",
+                "status",
+                "startedAt",
+                "updatedAt",
+                "basisToken",
+                "basis.selectedLeadListId",
+                "basis.workflowTableId",
+                "basis.reviewBatchRowHash or basis.reviewBatchRowIds",
+            ],
+            basisFields: [
+                "campaign revision or updatedAt",
+                "brief hash",
+                "selectedLeadListId",
+                "workflowTableId",
+                "bounded review-batch row ids/hash",
+                "filter choice",
+                "filter/rubric basis when present",
+            ],
+            compactOutputFields: [
+                "templateRecommendation",
+                "tokenFillRules",
+                "renderedSample",
+                "concerns",
+                "status",
+                "basisToken",
+                "outputAt",
+                "outputHash",
+                "error or retry detail",
+            ],
+        },
         usage: {
             codex: "After the user approves or auto-confirms the lead source, spawn both returned scout `name` values in one assistant turn only when the current Codex host exposes those custom agents.",
             claude: "After lead source approval, invoke both returned Task/Agent subagents in one assistant message only when the current Claude session lists those agents, so filter-leads and message generation run concurrently.",
-            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. Debug markdown/json artifacts are optional only. The message branch must load the full generate-messages prompt and may use the create-campaign-v2 message-review safety gate as a supplemental approval check. Join before message review.",
+            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. Debug markdown/json artifacts are optional only. The message branch must load the full generate-messages prompt, must read live campaign/review-batch state through scoped MCP/product tools, must reject mismatched selectedLeadListId/workflowTableId/campaign/workspace input, and may use the create-campaign-v2 message-review safety gate as a supplemental approval check. Join before message review.",
         },
     };
 }

package/dist/tools/sequencer.d.ts

@@ -35,6 +35,7 @@ export type AttachRecommendedSequenceInput = {
 export type AttachRecommendedSequenceResponse = {
     success: boolean;
     version: number;
+    currentStep?: string;
     template?: {
         name?: string;
         entryNodeId?: string;