@m13v/s4l 1.7.12-rc.1 → 1.7.12-rc.2

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -1,4 +1,4 @@
1
1
  {
2
- "version": "1.7.12-rc.1",
3
- "installedAt": "2026-08-17T21:32:52.343Z"
2
+ "version": "1.7.12-rc.2",
3
+ "installedAt": "2026-08-20T01:59:29.922Z"
4
4
  }
package/mcp/manifest.json CHANGED
@@ -2,7 +2,7 @@
2
2
  "dxt_version": "0.1",
3
3
  "name": "social-autoposter",
4
4
  "display_name": "S4L",
5
- "version": "1.7.12-rc.1",
5
+ "version": "1.7.12-rc.2",
6
6
  "description": "Draft, review, approve, and autopilot X/Twitter posts.",
7
7
  "long_description": "## **⚠️ The disclaimer above is generic Claude boilerplate.** Anthropic shows the same warning on every plugin regardless of what it does; any plugin has the same level of access as any app you download from the internet.\n\nS4L is an open source product developed by Mediar.ai Incorporated, a VC-backed San Francisco-based startup.\n\nTo get started:\n\n1\\. Copy this prompt: **Set me up on S4L plugin end to end**\n\n2\\. Quit with CMD+Q, reopen Claude, paste into a new chat.\n\nWhat happens next:\n\n* About every 5 minutes S4L scans X for posts that match your topics and drafts replies in your voice.\n* Drafts show up as review cards, usually the first within a few minutes. Nothing is posted automatically; you approve each one.\n* Posting autopilot stays off until you explicitly turn it on.",
8
8
  "author": {
package/mcp/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@m13v/s4l-mcp",
3
- "version": "1.7.12-rc.1",
3
+ "version": "1.7.12-rc.2",
4
4
  "private": true,
5
5
  "description": "Desktop MCP client for social-autoposter (X/Twitter rail): manual draft/review/approve loop, autopilot control, and stats. Thin wrapper over the existing pipeline scripts.",
6
6
  "license": "MIT",
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@m13v/s4l",
3
- "version": "1.7.12-rc.1",
3
+ "version": "1.7.12-rc.2",
4
4
  "description": "Automated social posting pipeline for Reddit, X/Twitter, LinkedIn, and Moltbook. Install as a Claude Code agent skill.",
5
5
  "bin": {
6
6
  "social-autoposter": "bin/cli.js",
@@ -840,8 +840,19 @@ def main() -> int:
840
840
 
841
841
  targets_by_pid: dict[int, dict] = {} # dedup across the two rules below
842
842
  for uuid, members in groups.items():
843
- if len(members) <= 1:
844
- continue # a healthy / interactive session — never touch.
843
+ if len(members) <= 1 and inflight is None:
844
+ # Queue unreadable: cannot tell a husk from a worker about to claim, so
845
+ # the conservative fallback never touches a lone session. When the queue
846
+ # IS readable, singleton groups get the same type-driven sweep as larger
847
+ # ones. The 2026-08-16 wedge: one warm session never exited, Desktop's
848
+ # per_task_limit therefore never spawned a sibling, and the old
849
+ # unconditional `len(members) <= 1: continue` guard meant the reaper and
850
+ # the scheduler starved each other for ~55 minutes (reapable=1, killed=0
851
+ # every cycle). Everything in `members` is already a metadata- or
852
+ # cwd-confirmed S4L worker (interactive sessions are spared upstream in
853
+ # snapshot()), so rule (1)'s invariant applies to a group of one exactly
854
+ # as it does to a group of ten: claimless past claim_grace = proven husk.
855
+ continue
845
856
  members.sort(key=lambda p: p["age"]) # ascending: newest first
846
857
 
847
858
  if inflight is not None: