fix(persona-chat): don't offer describe_photo in persona chats

describe_photo describes the file chat's photo. Persona chats have no photo,
so a vision-capable model was offered a tool that could only answer "No image
available". Persona chats now build their catalog with the vision gate off;
the thread tools still cover looking at images there.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

src/ai/persona_chat.rs

@@ -935,10 +935,14 @@ impl PersonaChatSession {
         // Build a no-photo tool catalog. `current_gate_opts_for_persona`
         // already probes each tool's backing table for presence and
         // honours the persona's `allow_agent_corrections` toggle.
-        let mut gate_opts = self.generator.current_gate_opts_for_persona(
-            backend.images_inline,
-            Some((user_id, &persona.persona_id)),
-        );
+        // `has_vision` only gates `describe_photo`, which describes the file
+        // chat's photo. A persona chat has no photo, so the tool could only
+        // ever answer "No image available" and waste a round. Vision in a
+        // persona chat goes through the thread tools (look_at_message_image,
+        // look_at_photo), which use the local vision model directly.
+        let mut gate_opts = self
+            .generator
+            .current_gate_opts_for_persona(false, Some((user_id, &persona.persona_id)));
         gate_opts.sms_thread_tools = row.anchor.is_some();
         let tools = InsightGenerator::build_tool_definitions(gate_opts);
         // Practice mode narrows the catalog to the thread itself — no