@sellable/install 0.1.89 → 0.1.91

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

@@ -17,14 +17,22 @@ Use the inherited Sellable MCP tools when available:
 - `search_signals` to find recent post lanes. Include `campaignOfferId` whenever
   the parent provides one so selected searches/lists stay attached to the
   campaign.
-- `fetch_post_engagers` to sample engagers from selected posts.
+- `select_promising_posts` to promote the exact posts you will sample into the
+  campaign UI before fetching engagers. In campaign-attached runs, do this
+  before the first `fetch_post_engagers` call so the user can see which posts
+  are being sampled.
+- `fetch_post_engagers` to sample engagers from promoted/selected posts.
 
 Process:
 
 1. Read the campaign brief, kickoff doc, or lane prompt supplied by the parent.
 2. Search 3-5 keyword/topic lanes, favoring fresh posts from the last 7-14 days.
-3. Select 3-5 promising posts when available.
-4. Fetch or sample engagers for selected posts and score rough ICP fit from visible headline/display-name cues only. Do not enrich people during viability estimation.
+3. Select 3-5 promising posts when available. If a `campaignOfferId` was
+   supplied, call `select_promising_posts({ campaignOfferId, selectionMode:
+"replace", selections, headlineICPCriteria, currentStep:
+"signal-discovery" })` before sampling so the watched Signal Discovery table
+   shows the promoted posts.
+4. 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.
 5. Estimate usable prospects per selected post from sampled pass rate. If the sample is good but volume is low, say how many more similar posts should be added or scraped.
 6. Return false positives and dead ends explicitly.
 

package/bin/sellable-install.mjs

@@ -444,6 +444,9 @@ function codexPluginMcp(opts) {
           type: "stdio",
           command,
           args,
+          env: {
+            SELLABLE_WATCH_MODE_DRIVER: "codex",
+          },
         },
       },
     };
@@ -455,6 +458,9 @@ function codexPluginMcp(opts) {
         type: "stdio",
         command: "npx",
         args: ["-y", opts.mcpPackage],
+        env: {
+          SELLABLE_WATCH_MODE_DRIVER: "codex",
+        },
       },
     },
   };
@@ -601,6 +607,12 @@ data, compare sources by source volume, sampled ICP fit, activity/warmth
 signals, cleanup risk, and confidence basis. If a user asks for a forecast,
 label it explicitly as not estimated from this run.
 
+For campaign-attached Signal Discovery sampling, promote/select the exact posts
+with \`select_promising_posts\` before \`fetch_post_engagers\` so the user can see
+which posts are being sampled in the watched app. The watch guide should say
+that we are pulling sample engagers from these posts to confirm the ICP is
+actually engaging and the source is viable.
+
 Every approval gate must include live campaign access after the readable inline
 content. Show a \`Watch link:\` line once the campaign shell exists. In normal
 customer runs, do not show local draft filenames or filesystem paths unless the
@@ -611,46 +623,42 @@ Never mention MCP namespaces, prompt chunking, plugin cache paths, missing
 linked skill versions, runbooks, or local skill files in normal customer-facing
 copy.
 
-## Codex Watch Browser Handoff
+## Codex Watch Link Handoff
 
-When a campaign tool returns \`watchUrl\`, treat it as an active browser handoff,
-not only a URL to print. A valid handoff link must be a safe direct
-\`/campaign-builder/{campaignId}?mode=claude\`
-URL. \`create_campaign.watchUrl\`, \`create_campaign({ campaignId }).watchUrl\`,
-and \`get_campaign.watchUrl\` are all acceptable only when they return that direct
-campaign-builder shape, with \`workspaceId\` used only as a safe routing hint when
-needed.
+When a campaign tool returns \`watchUrl\`, treat it as a user-opened app link, not
+as permission to drive the browser. A valid handoff link must be a direct
+\`/campaign-builder/{campaignId}?mode=claude|codex\` URL with the auth token and
+workspace routing needed for auto-login. \`create_campaign.watchUrl\`,
+\`create_campaign({ campaignId }).watchUrl\`, and \`get_campaign.watchUrl\` are all
+acceptable only when they return that direct campaign-builder shape.
 
-In Codex Desktop, when in-app browser control is available, open the returned watch link on the user's behalf. After opening it, inspect the browser-visible campaign state, then explain what the user is seeing before continuing with more campaign tools. Use customer language such as:
-
-\`\`\`text
-I’ll open the campaign view and keep it in sync as I build.
-\`\`\`
-
-If browser control is unavailable, provide the watch link without discussing the
-runtime limitation. In off-desktop Codex runs, make the direct watch link easy to
-copy/open, then continue with explicit customer-facing campaign progress and
-\`get_campaign_navigation_state\` orientation checks.
-Use this fallback shape:
+Never call browser-opening tools, shell \`open\`, Computer Use, or in-app browser
+automation just because a watch link exists. Print the link and tell the user to
+Command-Enter/click it when they want to watch in the app. Use this shape:
 
 \`\`\`text
 Watch link: {watchUrl}
 
-I’ll keep the campaign brief, lead source, and message-review steps explicit
-here as I build.
+Command-Enter or click that link to watch the campaign in Sellable. I’ll keep
+the brief, lead source, and message-review steps explicit here as I build.
 \`\`\`
 
-If opening the watch link lands on an auth, 404, permission, blank, or visible
-error state, report only what is visible, recover a fresh watch link once
-with \`create_campaign({ campaignId })\` or \`get_campaign\`, and try that link.
-Do not claim the browser was opened, inspected, or synchronized unless the
-visible browser state was actually observed. Do not claim the browser was opened
-unless it was opened and observed.
+The watch link should auto-login through the token in the URL. If the user says
+the link lands on auth, 404, permission, blank, or a visible error state, recover
+a fresh watch link once with \`create_campaign({ campaignId })\` or \`get_campaign\`
+and print that link. Do not claim the browser was opened, inspected, or
+synchronized.
 
 After every \`update_campaign({ campaignId, currentStep })\`, use
 \`get_campaign_navigation_state\` when available as a compact orientation check:
-match the browser-visible step to the saved campaign state, explain the current
-state in one sentence, and only then continue. Sender selection belongs at Settings after message approval and 10-row validation. After message validation, use Settings to help the user connect or select a LinkedIn sender. Explain Slack reply review before launch. After sender selection, attach the recommended sequence and move the watched UI to Send. Do not start the campaign or trigger a live send unless the user explicitly confirms that launch action outside UAT.
+match the saved campaign state to the expected watch-link step, explain the
+current state in one sentence, and only then continue. Sender selection belongs
+at Settings after message approval and 10-row validation. After message
+validation, use Settings to help the user connect or select a LinkedIn sender.
+Explain Slack reply review before launch. After sender selection, attach the
+recommended sequence and move the watched UI to Send. Do not start the campaign
+or trigger a live send unless the user explicitly confirms that launch action
+outside UAT.
 
 ## Names To Use
 
@@ -1516,6 +1524,10 @@ exec npx -y ${INSTALL_PACKAGE_SPEC} "$@"
   writeFile(binPath, shim, opts, 0o755);
 }
 
+function withWatchModeDriver(command, args, driver) {
+  return ["env", [`SELLABLE_WATCH_MODE_DRIVER=${driver}`, command, ...args]];
+}
+
 function mcpCommand(opts) {
   if (opts.server === "package") {
     return ["npx", ["-y", opts.mcpPackage]];
@@ -1534,6 +1546,12 @@ function mcpCommand(opts) {
   return ["hosted", [opts.hostedUrl]];
 }
 
+function mcpCommandForHost(opts, driver) {
+  const [command, args] = mcpCommand(opts);
+  if (command === "hosted") return [command, args];
+  return withWatchModeDriver(command, args, driver);
+}
+
 function installClaude(opts) {
   if (!commandExists("claude")) {
     const message =
@@ -1554,7 +1572,7 @@ function installClaude(opts) {
     patchClaudeAlwaysLoad(opts);
     return true;
   }
-  const [command, args] = mcpCommand(opts);
+  const [command, args] = mcpCommandForHost(opts, "claude");
   run("claude", ["mcp", "remove", "sellable"], {
     ...opts,
     dryRun: opts.dryRun,
@@ -1636,7 +1654,7 @@ function installCodex(opts) {
     const info = installCodexDesktopPlugin(opts);
     return { installed: true, ...info };
   }
-  const [command, args] = mcpCommand(opts);
+  const [command, args] = mcpCommandForHost(opts, "codex");
   run("codex", ["mcp", "remove", "sellable"], {
     ...opts,
     dryRun: opts.dryRun,

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@sellable/install",
-  "version": "0.1.89",
+  "version": "0.1.91",
   "type": "module",
   "description": "One-command installer for Sellable MCP in Claude Code and Codex",
   "bin": {

package/skill-templates/create-campaign.md

@@ -134,6 +134,12 @@ claim they ran in parallel and do not surface install status to the customer. In
 chat, call the downstream copy stage `message generation`;
 `message-validation.md` is only an internal proof artifact.
 
+For campaign-attached Signal Discovery sampling, promote/select the exact posts
+with `select_promising_posts` before `fetch_post_engagers` so the user can see
+which posts are being sampled in the watched app. The watch guide should say
+that we are pulling sample engagers from these posts to confirm the ICP is
+actually engaging and the source is viable.
+
 After find-leads returns a lead source and the user approves it, use the same
 registry pattern for the two post-lead branches. The create-campaign-v2 subskill
 calls `get_post_find_leads_scout_registry`, then launches the returned
@@ -159,45 +165,42 @@ linked skill versions, runbooks, npm/package details, repo-local files, VPS or
 browser automation limitations, or local skill files in normal customer-facing
 copy.
 
-## Codex Watch Browser Handoff
+## Codex Watch Link Handoff
 
-When a campaign tool returns `watchUrl`, treat it as an active browser handoff,
-not only a URL to print. A valid handoff link must be a safe direct
-`/campaign-builder/{campaignId}?mode=claude` URL. `create_campaign.watchUrl`,
+When a campaign tool returns `watchUrl`, treat it as a user-opened app link, not
+as permission to drive the browser. A valid handoff link must be a direct
+`/campaign-builder/{campaignId}?mode=claude|codex` URL with the auth token and
+workspace routing needed for auto-login. `create_campaign.watchUrl`,
 `create_campaign({ campaignId }).watchUrl`, and `get_campaign.watchUrl` are all
-acceptable only when they return that direct campaign-builder shape, with
-`workspaceId` used only as a safe routing hint when needed.
-
-In Codex Desktop, when in-app browser control is available, open the returned watch link on the user's behalf. After opening it, inspect the browser-visible campaign state, then explain what the user is seeing before continuing with more campaign tools. Use customer language such as:
-
-```text
-I’ll open the campaign view and keep it in sync as I build.
-```
+acceptable only when they return that direct campaign-builder shape.
 
-If browser control is unavailable, provide the watch link without discussing the
-runtime limitation. In off-desktop Codex runs, make the direct watch link easy
-to copy/open, then continue with explicit customer-facing campaign progress and
-`get_campaign_navigation_state` orientation checks.
-Use this fallback shape:
+Never call browser-opening tools, shell `open`, Computer Use, or in-app browser
+automation just because a watch link exists. Print the link and tell the user to
+Command-Enter/click it when they want to watch in the app. Use this shape:
 
 ```text
 Watch link: {watchUrl}
 
-I’ll keep the campaign brief, lead source, and message-review steps explicit
-here as I build.
+Command-Enter or click that link to watch the campaign in Sellable. I’ll keep
+the brief, lead source, and message-review steps explicit here as I build.
 ```
 
-If opening the watch link lands on an auth, 404, permission, blank, or visible
-error state, report only what is visible, recover a fresh watch link once with
-`create_campaign({ campaignId })` or `get_campaign`, and try that link. Do not
-claim the browser was opened, inspected, or synchronized unless the visible
-browser state was actually observed. Do not claim the browser was opened unless
-it was opened and observed.
+The watch link should auto-login through the token in the URL. If the user says
+the link lands on auth, 404, permission, blank, or a visible error state, recover
+a fresh watch link once with `create_campaign({ campaignId })` or `get_campaign`
+and print that link. Do not claim the browser was opened, inspected, or
+synchronized.
 
 After every `update_campaign({ campaignId, currentStep })`, use
 `get_campaign_navigation_state` when available as a compact orientation check:
-match the browser-visible step to the saved campaign state, explain the current
-state in one sentence, and only then continue. Sender selection belongs at Settings after message approval and 10-row validation. After message validation, use Settings to help the user connect or select a LinkedIn sender. Explain Slack reply review before launch. After sender selection, attach the recommended sequence and move the watched UI to Send. Do not start the campaign or trigger a live send unless the user explicitly confirms that launch action outside UAT.
+match the saved campaign state to the expected watch-link step, explain the
+current state in one sentence, and only then continue. Sender selection belongs
+at Settings after message approval and 10-row validation. After message
+validation, use Settings to help the user connect or select a LinkedIn sender.
+Explain Slack reply review before launch. After sender selection, attach the
+recommended sequence and move the watched UI to Send. Do not start the campaign
+or trigger a live send unless the user explicitly confirms that launch action
+outside UAT.
 
 ## Names To Use