social-autoposter 1.6.50 → 1.6.52

package/mcp/dist/index.js

@@ -309,10 +309,24 @@ function renderDraftsTable(plan) {
         const author = c.thread_author ? `@${c.thread_author}` : "(unknown thread)";
         const style = c.engagement_style ?? "?";
         const reply = c.reply_text ?? "(empty)";
-        const link = c.link_url ? `  ·  link: ${c.link_url}` : "";
-        return (`[${n}] ${author}  (style: ${style})${link}\n` +
-            `    ${reply.replace(/\n/g, "\n    ")}\n` +
-            `    thread: ${c.candidate_url ?? "?"}`);
+        // The literal tail URL is NOT known yet: at post time a short link is minted
+        // from this target (e.g. fazm.ai/cc -> s4l.ai/r/<code>). Approved drafts
+        // always carry the link (post_drafts forces TWITTER_TAIL_LINK_RATE=1.0), so
+        // this is the target that WILL be appended. Show the TARGET only; never
+        // pre-mint the real /r/ code (that would waste pool codes / split clicks).
+        const link = c.link_url
+            ? `\n    + link (appended as a short link at post time): ${c.link_url}`
+            : "";
+        // The original tweet we're replying to — context the reviewer needs to judge
+        // the draft. Already in the plan; just surface it.
+        const threadText = c.thread_text
+            ? `\n    in reply to: ${c.thread_text.replace(/\s+/g, " ").trim().slice(0, 280)}`
+            : "";
+        return (`[${n}] ${author}  (style: ${style})` +
+            `${threadText}\n` +
+            `    draft: ${reply.replace(/\n/g, "\n    ")}` +
+            `${link}\n` +
+            `    thread url: ${c.candidate_url ?? "?"}`);
     })
         .join("\n\n");
 }
@@ -331,6 +345,16 @@ async function postApproved(batchId, plan) {
         timeoutMs: 900_000,
         env: {
             SAPS_SKIP_CAMPAIGN_SUFFIX: "1",
+            // Manual approval is an EXCEPTION to the tail-link A/B. The cron pipeline
+            // runs TWITTER_TAIL_LINK_RATE=0.9 (from .env) so ~10% of autopilot posts
+            // ship link-less as an experiment arm. But when the user hand-reviews a
+            // draft, sees the link target in the table, and approves it, dropping the
+            // link is surprising and unwanted. Force 1.0 here so every approved draft
+            // carries its link. This wins over .env / process.env because run() spreads
+            // opts.env AFTER process.env, and twitter_post_plan.py never load_dotenv's
+            // with override, so nothing clobbers it. Cron is untouched (it never goes
+            // through this MCP path), so the 0.9 experiment keeps running there.
+            TWITTER_TAIL_LINK_RATE: "1.0",
             // The poster attaches to the twitter-harness Chrome over CDP. The cron
             // pipeline exports this from skill/lib/twitter-backend.sh; the MCP path
             // must set it explicitly or twitter_browser.py fails with "No twitter-
@@ -582,7 +606,9 @@ server.registerTool("draft_cycle", {
     description: "Scan X and draft replies on this machine, then return ALL drafts as a numbered table " +
         "for review. This tool POSTS NOTHING. Show the table to the user and ask which numbers " +
         "to post and which to rewrite, then call `post_drafts` with their decision and the " +
-        "returned batch_id. Flow: discover -> draft -> review in chat -> post_drafts.",
+        "returned batch_id. The table MUST show, per draft: the thread being replied to " +
+        "(thread_text), the draft reply, and the link target (link_url) when present; never " +
+        "drop those columns. Flow: discover -> draft -> review in chat -> post_drafts.",
     inputSchema: {
         project: z
             .string()
@@ -639,11 +665,15 @@ server.registerTool("draft_cycle", {
     const message = `Drafted ${count} ${count === 1 ? "reply" : "replies"} for "${proj}" ` +
         `(batch ${drafted.batchId}). NOTHING has been posted yet.\n\n` +
         `${table}\n\n` +
-        `Show this list to the user and ask which to post and which to edit. They can reply ` +
-        `however is natural, e.g. "post 1, 3 and 5", "edit 2: <new wording>", "post all", or ` +
-        `"skip all". Editing a draft also posts it. Then call the post_drafts tool with ` +
-        `batch_id "${drafted.batchId}" and their decision (post: [numbers], edits: [{n, text}], ` +
-        `or post_all: true). Do not post anything the user didn't ask for.`;
+        `Show this list to the user and ask which to post and which to edit. When you render ` +
+        `the table, ALWAYS include, for every draft: the thread it replies to (in reply to / ` +
+        `thread_text), the draft text, and the link target (link_url) if present. Do NOT drop ` +
+        `the thread or link columns. Note the literal short link is appended at post time and ` +
+        `an A/B gate may omit it, so present link_url as the target, not a guaranteed final URL. ` +
+        `They can reply however is natural, e.g. "post 1, 3 and 5", "edit 2: <new wording>", ` +
+        `"post all", or "skip all". Editing a draft also posts it. Then call the post_drafts ` +
+        `tool with batch_id "${drafted.batchId}" and their decision (post: [numbers], ` +
+        `edits: [{n, text}], or post_all: true). Do not post anything the user didn't ask for.`;
     return {
         content: [{ type: "text", text: message }],
         structuredContent: {
@@ -659,7 +689,15 @@ server.registerTool("draft_cycle", {
                 n: i + 1,
                 author: c.thread_author,
                 tweet_url: c.candidate_url,
+                // The original tweet being replied to — reviewer context. Hosts that
+                // surface ONLY structuredContent (Claude Desktop, Fazm Code tab) need
+                // this here or the relayed table loses the thread it's responding to.
+                thread_text: c.thread_text,
                 reply_text: c.reply_text,
+                // Target link only. The literal /r/<code> short link is minted at post
+                // time and an A/B gate may omit it; do not pre-mint here.
+                link_url: c.link_url,
+                style: c.engagement_style,
                 language: c.language,
             })),
         },

package/mcp-servers/browser-harness/server.py

@@ -249,11 +249,32 @@ def _build_chrome_cmd() -> list[str]:
         cmd.append(f"--window-position={win_pos}")
         cmd.append(f"--window-size={win_size}")
 
-    # Open a real tab so CDP has something to attach to immediately.
-    cmd.append("about:blank")
+    # Open a REAL http(s) page (NOT about:blank) so the harness daemon's
+    # attach_first_page() finds an existing real tab and reuses it. Upstream
+    # browser-harness creates a throwaway about:blank on every re-attach where
+    # is_real_page() is false (which about:blank is); launching blank feeds that
+    # loop, piling up orphan tabs that cleanup_harness_tabs then has to sweep.
+    # Landing URL is derived from the profile's platform; BH_LAUNCH_URL overrides.
+    cmd.append(_launch_url())
     return cmd
 
 
+def _launch_url() -> str:
+    """Initial tab URL for the managed Chrome. MUST be a real http(s) page, not
+    about:blank, so the browser-harness daemon reuses it instead of spawning
+    throwaway blank tabs (see _build_chrome_cmd). Derived from PROFILE_NAME;
+    BH_LAUNCH_URL overrides for non-standard profiles."""
+    override = os.environ.get("BH_LAUNCH_URL", "").strip()
+    if override:
+        return override
+    name = PROFILE_NAME.lower()
+    if "linkedin" in name:
+        return "https://www.linkedin.com/feed/"
+    if "reddit" in name:
+        return "https://www.reddit.com/"
+    return "https://x.com"
+
+
 def ensure_chrome() -> dict:
     """Make sure our managed Chrome is running on PORT. Idempotent."""
     if _cdp_alive():

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "social-autoposter",
-  "version": "1.6.50",
+  "version": "1.6.52",
   "description": "Automated social posting pipeline for Reddit, X/Twitter, LinkedIn, and Moltbook. Install as a Claude Code agent skill.",
   "bin": {
     "social-autoposter": "bin/cli.js"

package/scripts/engagement_styles.py

@@ -182,20 +182,24 @@ STYLES = {
     },
     "ig_walkin_storefront_playbook": {
         "description": (
-            "Door-to-door SMB story for mk0r: 'i was [working-class role]. i "
-            "drove past N businesses with no website every day. a friend told "
-            "me about mk0r. i walked in / sat at the counter. i opened mk0r.com, "
-            "typed one prompt, the site built itself while [owner] watched. "
-            "they paid me $[300-500] cash from a [coffee tin / register]. N "
-            "months later i've signed [N] of these = $[N]/mo recurring.' Ends "
-            "with the mk0r.com footer."
+            "Product demo for mk0r: show mk0r building a REAL website for a "
+            "local business that has none. 'i noticed [a local spot] near me "
+            "with no website, just a maps pin and one blurry photo. i opened "
+            "mk0r.com, described the place in one prompt, and watched it build "
+            "a real site, hero, services, hours, a call button, live in "
+            "minutes.' Focus on the CAPABILITY and the speed, and what it "
+            "means for a small business to finally have a real site online. NO "
+            "earnings, NO 'they paid me', NO recurring-revenue or 'signed N "
+            "clients' totals. Reference mk0r.com plainly."
         ),
         "example": (
-            "i was 27. shipping clerk at a parts warehouse outside fresno for "
-            "4 years, $21 an hour. i drove past 14 auto shops on the way home "
-            "every night. ... i opened mk0r.com on his waiting room table and "
-            "turned it around. he paid me from the register drawer. ... 23 "
-            "auto shops signed since march. 🔗 mk0r.com"
+            "there's a tire shop two blocks from me thats been open fifteen "
+            "years. i went to send a friend the link and there was no link, "
+            "just a maps pin and one blurry photo someone else uploaded. so i "
+            "opened mk0r.com and described the shop in one prompt, brakes, "
+            "tires, the hours, the phone number. it built the whole site while "
+            "i watched, a hero, a services list, a map, one big call button. "
+            "fifteen years open and it finally has a front door online. mk0r.com"
         ),
         "best_in": {
             "instagram": [
@@ -204,29 +208,40 @@ STYLES = {
             ],
         },
         "note": (
-            "Caption is the niche walk-in arc (working-class persona, drive-by "
-            "niche, first walk-in success, recurring revenue total). Overlay "
-            "is the 5-beat playbook (title card + 4 step cards). MUST include "
-            "'mk0r.com' in caption and the mk0r.com footer. project_name='mk0r' "
-            "on the row. Fires when TARGET=product AND selected_project=mk0r."
+            "Caption is the product-demo arc: a real local business with no "
+            "website, mk0r builds a real site from one prompt, what it means "
+            "for the owner to finally be online. NO income/earnings/'they paid "
+            "me'/recurring-revenue/'signed N clients' framing -- that arc "
+            "tripped a Meta fraud-and-deceptive-practices restriction on "
+            "2026-06-02 (matt_diak link-sharing restricted 30d). Reference "
+            "'mk0r.com' plainly in the caption. project_name='mk0r' on the "
+            "row. Fires when TARGET=product AND selected_project=mk0r."
         ),
         "target_chars": 1800,
     },
     "ig_studyly_failing_student_arc": {
         "description": (
-            "Failing-student outcome arc for studyly: 'i was [age], [program]. "
-            "i [failed/scored low] on [exam]. i [reread/flashcards/highlights] "
-            "for weeks. nothing worked. a friend sent me studyly.io. i pasted "
-            "my [notes/chapter] in. it quizzed me until i could answer without "
-            "looking. i got [higher score].' Closes with rereading-is-theater "
-            "lesson + studyly.io footer."
+            "Study-method arc for studyly: 'i was [age], [program]. i was "
+            "[rereading/highlighting/making flashcards] for hours and still "
+            "blanking the moment the wording changed. a friend sent me "
+            "studyly.io. i pasted my [notes/chapter] in and it quizzed me on "
+            "my own material until i could answer without looking, rewording "
+            "each question so i couldnt pattern-match. for the first time i "
+            "could tell what i actually knew from what i only recognized.' "
+            "Focus on the METHOD shift (active recall vs rereading, testing on "
+            "your own deck). Closes with the rereading-is-theater lesson + "
+            "studyly.io footer. Do NOT promise or fabricate specific "
+            "before/after exam scores as a typical result."
         ),
         "example": (
-            "i was 19. premed track, third semester, organic chemistry. i "
-            "failed the first orgo exam. 47 out of 100. ... i pasted my notes "
-            "into studyly.io at 2am. ... i got 78 on the second exam. ... the "
-            "lesson is rereading is recognizing. close the book. let something "
-            "ask you. studyly.io"
+            "i was 19. premed, third semester, organic chemistry. i had reread "
+            "the chapter four times and could still blank on a mechanism the "
+            "second the wording changed. i pasted my notes into studyly.io at "
+            "2am and it asked me the things i thought i knew, rewording each "
+            "one so i couldnt coast on the first three words. that was the "
+            "first night i could actually tell what i knew from what i only "
+            "recognized. the lesson is rereading is recognizing. close the "
+            "book. let something ask you. studyly.io"
         ),
         "best_in": {
             "instagram": [
@@ -237,8 +252,13 @@ STYLES = {
         "note": (
             "Caption is shorter than mk0r (1400-1900 chars). 'here is a story.' "
             "opener optional. MUST include 'studyly.io' footer. "
-            "project_name='studyly'. Lesson is always rereading-is-theater. "
-            "Fires when TARGET=product AND selected_project=studyly."
+            "project_name='studyly'. Lesson is always rereading-is-theater. NO "
+            "specific before/after exam scores or 'failed -> passed/topped the "
+            "class' miracle jumps presented as a typical/guaranteed outcome -- "
+            "exaggerated-results claims are a deceptive-practices signal (same "
+            "Meta rail that restricted mk0r 2026-06-02); keep outcomes "
+            "qualitative and personal. Fires when TARGET=product AND "
+            "selected_project=studyly."
         ),
         "target_chars": 1650,
     },

package/scripts/ig_post_type_picker.py

@@ -159,6 +159,21 @@ def main():
         # product->organic fallback this preserves cadence without weakening
         # the cooldown.
         other = "product" if target == "organic" else "organic"
+        # ...but ONLY if the other type is still eligible (config weight > 0).
+        # A type explicitly disabled via post_type_weights=0 (e.g. product
+        # paused during an IG link-sharing restriction) must NEVER be posted
+        # through the empty-queue fallback, otherwise weight=0 is not a real
+        # off switch. If the only eligible type's queue is empty, exit 2
+        # (nothing to post this fire) rather than leak the disabled type.
+        if other not in eligible:
+            sys.stderr.write(
+                f"queue empty for eligible type '{target}'; fallback type "
+                f"'{other}' is disabled (config_weights={type_weights_cfg}), "
+                "not leaking it"
+                + (f" (target_account={account})" if account else "")
+                + "\n"
+            )
+            sys.exit(2)
         if other == "product":
             row = _pick_product_with_cooldown()
         else: