personas: elevate to server with per-persona fact scoping

Move personas off the mobile client into ImageApi as first-class
records, and scope entity_facts by persona so each one builds its own
voice over a shared entity graph. The new include_all_memories flag
lets a persona opt back into the full hive-mind pool for human
browsing of /knowledge/*; agentic generation always stays in-voice.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

src/ai/insight_chat.rs

@@ -50,6 +50,10 @@ pub struct ChatTurnRequest {
     /// In amend mode, persisted into the new insight row's system message.
     /// None / empty = no change.
     pub system_prompt: Option<String>,
+    /// Active persona id for this turn. Tools that write to
+    /// `entity_facts` tag the new rows with it; `recall_facts_for_photo`
+    /// scopes its read to it. None defaults to `"default"`.
+    pub persona_id: Option<String>,
     /// When true, write a new insight row (regenerating title) instead of
     /// updating training_messages on the existing row.
     pub amend: bool,
@@ -231,6 +235,13 @@ impl InsightChatService {
             bail!("user_message exceeds 8192 chars");
         }
 
+        let active_persona = req
+            .persona_id
+            .clone()
+            .filter(|s| !s.trim().is_empty())
+            .unwrap_or_else(|| "default".to_string());
+        span.set_attribute(KeyValue::new("persona_id", active_persona.clone()));
+
         let normalized = normalize_path(&req.file_path);
 
         // 1. Acquire the per-(library, file) async mutex. Two concurrent
@@ -464,6 +475,7 @@ impl InsightChatService {
                             &ollama_client,
                             &image_base64,
                             &normalized,
+                            &active_persona,
                             &loop_cx,
                         )
                         .await;
@@ -737,6 +749,11 @@ impl InsightChatService {
         insight: crate::database::models::PhotoInsight,
         tx: tokio::sync::mpsc::Sender<ChatStreamEvent>,
     ) -> Result<()> {
+        let active_persona = req
+            .persona_id
+            .clone()
+            .filter(|s| !s.trim().is_empty())
+            .unwrap_or_else(|| "default".to_string());
         let raw_history = insight.training_messages.as_ref().ok_or_else(|| {
             anyhow!("insight has no chat history; regenerate this insight in agentic mode")
         })?;
@@ -826,6 +843,7 @@ impl InsightChatService {
                 tools,
                 &image_base64,
                 &normalized,
+                &active_persona,
                 max_iterations,
                 &tx,
             )
@@ -915,6 +933,11 @@ impl InsightChatService {
         normalized: String,
         tx: tokio::sync::mpsc::Sender<ChatStreamEvent>,
     ) -> Result<()> {
+        let active_persona = req
+            .persona_id
+            .clone()
+            .filter(|s| !s.trim().is_empty())
+            .unwrap_or_else(|| "default".to_string());
         let effective_backend = resolve_bootstrap_backend(req.backend.as_deref())?;
         let is_hybrid = effective_backend == "hybrid";
 
@@ -1008,6 +1031,7 @@ impl InsightChatService {
                 tools,
                 &image_base64,
                 &normalized,
+                &active_persona,
                 max_iterations,
                 &tx,
             )
@@ -1157,6 +1181,7 @@ impl InsightChatService {
         tools: Vec<Tool>,
         image_base64: &Option<String>,
         normalized: &str,
+        active_persona: &str,
         max_iterations: usize,
         tx: &tokio::sync::mpsc::Sender<ChatStreamEvent>,
     ) -> Result<AgenticLoopOutcome> {
@@ -1235,6 +1260,7 @@ impl InsightChatService {
                             ollama_client,
                             image_base64,
                             normalized,
+                            active_persona,
                             &cx,
                         )
                         .await;