From 2e1c0a9a68903d1726f4b87a6ca6b4c581c8d60e Mon Sep 17 00:00:00 2001
From: "Rodrigo Rodriguez (Pragmatismo)" <me@rodrigorodriguez.com>
Date: Fri, 24 Oct 2025 23:36:16 -0300
Subject: [PATCH] Add caching mechanism to BotOrchestrator for user input
 responses

---
 src/bot/mod.rs | 47 ++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 40 insertions(+), 7 deletions(-)

diff --git a/src/bot/mod.rs b/src/bot/mod.rs
index 3f4d4ae0c..73bbed0c2 100644
--- a/src/bot/mod.rs
+++ b/src/bot/mod.rs
@@ -9,16 +9,24 @@ use serde_json;
 use std::collections::HashMap;
 use std::sync::Arc;
 use tokio::sync::mpsc;
+use tokio::sync::Mutex;
 use uuid::Uuid;
+use redis::AsyncCommands;
+use reqwest::Client;
 
 pub struct BotOrchestrator {
     pub state: Arc<AppState>,
+    pub cache: Arc<Mutex<std::collections::HashMap<String, String>>>,
 }
 
 impl BotOrchestrator {
-    pub fn new(state: Arc<AppState>) -> Self {
-        Self { state }
+pub fn new(state: Arc<AppState>) -> Self {
+    Self {
+        state,
+        cache: Arc::new(Mutex::new(std::collections::HashMap::new())),
     }
+}
+}
 
     pub async fn handle_user_input(
         &self,
@@ -293,15 +301,39 @@ impl BotOrchestrator {
             session_manager.get_conversation_history(session.id, session.user_id)?
         };
 
-        for (role, content) in history {
+        // Prompt compactor: keep only last 10 entries to limit size
+        let recent_history = if history.len() > 10 {
+            &history[history.len() - 10..]
+        } else {
+            &history[..]
+        };
+
+        for (role, content) in recent_history {
             prompt.push_str(&format!("{}: {}\n", role, content));
         }
         prompt.push_str(&format!("User: {}\nAssistant:", message.content));
 
-        self.state
+        // Check in-memory cache for existing response
+        {
+            let cache = self.cache.lock().await;
+            if let Some(cached) = cache.get(&prompt) {
+                return Ok(cached.clone());
+            }
+        }
+
+        // Generate response via LLM provider
+        let response = self.state
             .llm_provider
             .generate(&prompt, &serde_json::Value::Null)
-            .await
+            .await?;
+
+        // Store the new response in cache
+        {
+            let mut cache = self.cache.lock().await;
+            cache.insert(prompt.clone(), response.clone());
+        }
+
+        Ok(response)
     }
 
     pub async fn stream_response(
@@ -447,12 +479,12 @@ impl BotOrchestrator {
             }
 
             analysis_buffer.push_str(&chunk);
-            if analysis_buffer.contains("<|channel|>") && !in_analysis {
+            if analysis_buffer.contains("**") && !in_analysis {
                 in_analysis = true;
             }
 
             if in_analysis {
-                if analysis_buffer.ends_with("final<|message|>") {
+                if analysis_buffer.ends_with("final") {
                     debug!(
                         "Analysis section completed, buffer length: {}",
                         analysis_buffer.len()
@@ -695,6 +727,7 @@ impl Default for BotOrchestrator {
     fn default() -> Self {
         Self {
             state: Arc::new(AppState::default()),
+            cache: Arc::new(Mutex::new(std::collections::HashMap::new())),
         }
     }
 }