osborn 0.9.100 → 0.9.101

package/.claude/skills/browser-screen-recorder/SKILL.md

@@ -13,12 +13,21 @@ capture, DevTools diagnostics, and metrics. It can speak into the page's mic
 (synthesized on the fly), hear and transcribe what it plays back, click via a
 natural-language brain, and improvise conversations. Works on any website.
 
-USE IT BROADLY — not just for tests: MANDATORY for debugging anything
+USE IT BROADLY — not just for tests: it is THE QA PROTOCOL for UI work —
+IMMEDIATELY AFTER building/styling/changing any frontend, the final build
+step is recording proof at mobile + desktop widths (never ask the user to
+eyeball or screenshot your own change). Also MANDATORY for debugging anything
 web-reachable (real browser + DevTools on camera + /logs, never debug a
 frontend blind) and for verifying deployed web changes (journey + clip, or it
 is a claim, not a verification). Also: reproducing user issues (mobile view),
-the meeting-canvas cast source, site research, voice testing. Every use
-returns media — review it and deliver it to the user, no silent runs.
+the meeting-canvas cast source, site research, voice testing.
+
+IRON LAW (applies even before you fetch the served copy): NO RESULT REPORT
+WITHOUT FRAME-REVIEW EVIDENCE — a green assertion is a claim; a read frame is
+proof. The served skill contains a Run Checklist you MUST copy into your
+response and check off; `scripts/review-run.sh <clip>` in the harness does the
+frame extraction. Media is reviewed by you AND delivered to the user, every
+run, no exceptions.
 
 ## UPDATE CHECK (notify-first, run before major runs)
 This packaged copy ships with every machine and may lag the served version.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "osborn",
-  "version": "0.9.100",
+  "version": "0.9.101",
   "description": "Voice AI coding assistant - local agent that connects to Osborn frontend",
   "type": "module",
   "bin": {