chore(newos4): commit accumulated maintenance outputHEADv4.6.156main

The daily collector and the feedback pipeline had left 57 modified files uncommitted (skills,
modules-index, CHANGELOG, feedback decisions, debug counters). The public mirror publishes git
HEAD, so without this the published repo stayed behind the tree - and it blocked the publisher
outright once uncommitted subtrees started failing loudly.

1 files changed, 15 insertions, 0 deletions

diff --git a/skills/emergency-rollback.json b/skills/emergency-rollback.json
new file mode 100644
index 0000000..620521f
--- /dev/null
+++ b/skills/emergency-rollback.json
@@ -0,0 +1,15 @@
+{
+  "name": "emergency-rollback",
+  "description": "Quick recovery skill for when module edits break the system. Restore from checkpoints and revert bad changes.",
+  "tags": [
+    "emergency",
+    "rollback",
+    "recovery",
+    "restore",
+    "module-edit"
+  ],
+  "category": "system",
+  "created": "2026-10-16T16:28:11.588Z",
+  "updated": "2026-10-16T16:30:38.270Z",
+  "content": "---\nname: emergency-rollback\ndescription: Quick recovery skill for when module edits break the system. Restore from checkpoints and revert bad changes.\ntags: [emergency, rollback, recovery, restore, module-edit]\ncategory: system\n---\n## Steps\n1. **Assess the emergency:**\n   - get_status — check FS, modules, memory health\n   - get_module_status — identify which module is broken\n   - health_check — run full system diagnostic\n\n2. **Check for checkpoints:**\n   - checkpoint-list — find pre-edit snapshots\n   - Prioritize: most recent, pre-module-edit checkpoints\n\n3. **Rollback options (in order of preference):**\n\n   **Option A: VFS Rollback (if checkpoint exists)**\n   - rollback <checkpoint-name> — restore entire VFS state\n   - Verify: get_status again\n\n   **Option B: Module Revert (edit broke specific module)**\n   - diff_module <ModuleName> — see what changed\n   - reload_module <ModuleName> — pull original from server\n   - If reload fails: view_module <Name> → copy original code manually\n\n   **Option C: Ship Module Rollback (if ship_module failed)**\n   - Auto-rollback happens if ship_module fails tests\n   - Check: list_modified_modules — find any pending bad edits\n\n   **Option D: Full System Restore**\n   - If checkpoint exists: rollback <name>\n   - If no checkpoint: restore from last known good module versions\n\n4. **Post-recovery verification:**\n   - health_check — run full diagnostic\n   - Test critical commands: help, status, ls\n   - Verify agents still alive: list_agents\n\n5. **Create recovery checkpoint:**\n   - After stabilization: checkpoint emergency-recovery\n   - Document what went wrong in session_save\n\n## Critical commands:\n- get_module_status → see modified modules\n- diff_module <Name> → see changes\n- reload_module <Name> → restore original\n- rollback <name> → VFS restore\n- health_check → verify recovery\n\n## Prevention tips:\n- Always checkpoint before risky edits\n- Use ship_module (auto-rollback on failure)\n- Test with test_module before committing\n- Keep at least one pre-edit checkpoint\n\n## When to use:\n- Module edit causes errors\n- Commands not registering\n- System unresponsive or broken\n- Any \"the system is broken\" report\n"
+}
\ No newline at end of file