fluxflow-cli 3.19.1 → 3.19.2

package/ARCHITECTURE.md

@@ -1,3 +1,6 @@
+##### **THIS DOCUMENTATION FILE IS OUTDATED & NO LONGER MAINTAINED. TO SEE LATEST ARCHITECTURAL DOCUMENTATION & CHANGES VISIT: [FluxFlow Architecture Documentation](https://fluxflow-cli.onrender.com/architecture/)**
+##### [Changelog](https://fluxflow-cli.onrender.com/changelog.html)
+
 # 🏛️ Architecture & Design
 
 Flux Flow is built on a modern, reactive stack that brings web-like development paradigms to the terminal. It utilizes a custom agentic loop for reasoning and a unique dual-model system for background processing.

package/README.md

@@ -11,7 +11,7 @@
 
 📖 **[Official Documentation](https://fluxflow-cli.onrender.com/)**
 
-**Flux Flow** is not just another CLI—it's a high-speed, sassy, and goal-oriented CLI AI Agent powered by the latest Gemini/Gemma frontier models. Designed for developers who demand a premium UI/UX while managing complex file-system tasks, web research, and autonomous workflows.
+**Flux Flow** is not just another CLI—it's a high-speed, sassy, and goal-oriented CLI AI Agent powered by multi-provider frontier models (Google Gemini, DeepSeek, OpenRouter, and more). Designed for developers who demand a premium UI/UX while managing complex file-system tasks, web research, and autonomous workflows.
 
 ---
 

package/TOOLS.md

@@ -1,3 +1,6 @@
+##### **THIS DOCUMENTATION FILE IS OUTDATED & NO LONGER MAINTAINED. TO SEE LATEST TOOLS DOCUMENTATION & AVAILABILITY VISIT: [FluxFlow Tools Documentation](https://fluxflow-cli.onrender.com/tools/)**
+##### [Changelog](https://fluxflow-cli.onrender.com/changelog.html)
+
 # 🧰 Agent Tools & Capabilities
 
 Flux Flow provides a robust set of tools that allow the AI to interact with the file system, execute code, and search the web. The availability of these tools depends on the active operating mode.

package/UI_FEATURES.md

@@ -1,3 +1,6 @@
+##### **THIS DOCUMENTATION FILE IS OUTDATED & NO LONGER MAINTAINED. TO SEE LATEST UI FEATURES & CHANGES VISIT: [FluxFlow Documentation](https://fluxflow-cli.onrender.com/)**
+##### [Changelog](https://fluxflow-cli.onrender.com/changelog.html)
+
 # 🎮 User Interface & Interaction Features
 
 Flux Flow is designed to be a high-performance terminal application. Beyond basic chat, it includes a variety of advanced UI features and human-in-the-loop controls to keep you in command of the agent.
@@ -88,4 +91,4 @@ Flux Flow is a "Self-Healing" agent. It actively monitors its own environment to
 
 - **Startup Integrity Check**: On launch, Flux performs a "Heartbeat Check" on its internal engines.
 - **Automatic Recovery**: If a dependency is missing, you will see a `🔧 [SYSTEM] Initializing...` message. Flux will autonomously download and configure the required binaries using `pnpm` or `npx` fallbacks, keeping you informed every step of the way.
-- **Silent Maintenance**: Once the engine is ready, you'll receive a `✅ [SYSTEM] All dependencies installed successfully.` confirmation.
+- **Silent Maintenance**: Once the engine is ready, you'll receive a `✅ [SYSTEM] All dependencies installed successfully.` confirmation.

package/dist/fluxflow.js

@@ -6977,7 +6977,8 @@ TOOL RULES:
 ${mode === "Flux" ? `- JSON ESCAPE ALL LITERAL ESCAPE SEQUENCES IN TOOL ARGUMENTS
 - SAME file, MULTIPLE edits? ONE PatchFile (\u226415 blocks) \u2190 PRIORITY
 - Tool denied? Ask for guidance \u2190 MANDATORY
-- Need text or huge files? SearchKeyword > Full Read
+- Need text or HUGE file? SearchKeyword > Full Read
+- MUST AVOID UNNECESSARY LARGE-FILE CHUNK READS
 ` : ""}
 - COMMUNICATION WITH USER -
 - [tool:functions.Ask(question="...", optionA="title::description", ...MAX4)]. Ambiguity: MUST for path divergence, security risk. Ask, don't finish/guess. Keep titles short
@@ -9863,7 +9864,7 @@ ${projectContextBlock}${isMemoryEnabled ? `
 -- MEMORY RULES --
 - Subtly Personalize with  RELEVENT CONTEXTUAL MEMORIES. Auto Saves
 ` : ""}
--- SECURITY RULES --
+-- SECURITY POLICIES --
 - Sensitive files? Ask before Read${isSystemDir ? "\n- PROTECTED DIRECTORY" : ""}
 
 -- CHAT FORMATTING --
@@ -15372,34 +15373,51 @@ var init_ai = __esm({
         for (const line of lines) {
           const cleanLine = line.trim();
           if (!cleanLine || !cleanLine.startsWith("data: ")) continue;
-          if (cleanLine === "data: [DONE]") break;
-          try {
-            const json = JSON.parse(cleanLine.substring(6));
-            const delta = json.choices?.[0]?.delta;
-            const usage = json.usage;
-            if (usage) {
-              latestUsageMetadata = {
-                totalTokenCount: usage.total_tokens || usage.prompt_tokens + usage.completion_tokens,
-                promptTokenCount: usage.prompt_tokens || 0,
-                candidatesTokenCount: usage.completion_tokens || 0,
-                cachedContentTokenCount: usage.prompt_tokens_details?.cached_tokens || 0,
-                thoughtsTokenCount: usage.completion_tokens_details?.reasoning_tokens || 0
-              };
-              hasNewData = true;
-            }
-            if (delta) {
-              const thought = delta.reasoning_content || null;
-              if (thought) {
-                pendingParts.push({ text: thought, thought: true });
-                hasNewData = true;
+          let isDone = false;
+          if (cleanLine === "data: [DONE]") {
+            isDone = true;
+          } else {
+            try {
+              const json = JSON.parse(cleanLine.substring(6));
+              const delta = json.choices?.[0]?.delta;
+              const usage = json.usage;
+              if (json.choices?.[0]?.finish_reason) {
+                isDone = true;
               }
-              if (delta.content) {
-                pendingParts.push({ text: delta.content });
+              if (usage) {
+                latestUsageMetadata = {
+                  totalTokenCount: usage.total_tokens || usage.prompt_tokens + usage.completion_tokens,
+                  promptTokenCount: usage.prompt_tokens || 0,
+                  candidatesTokenCount: usage.completion_tokens || 0,
+                  cachedContentTokenCount: usage.prompt_tokens_details?.cached_tokens || 0,
+                  thoughtsTokenCount: usage.completion_tokens_details?.reasoning_tokens || 0
+                };
                 hasNewData = true;
               }
+              if (delta) {
+                const thought = delta.reasoning_content || null;
+                if (thought) {
+                  pendingParts.push({ text: thought, thought: true });
+                  hasNewData = true;
+                }
+                if (delta.content) {
+                  pendingParts.push({ text: delta.content });
+                  hasNewData = true;
+                }
+              }
+            } catch (e) {
             }
-          } catch (e) {
           }
+          if ((isDone || Date.now() - lastFlushTime >= 150) && hasNewData) {
+            yield {
+              candidates: pendingParts.length > 0 ? [{ content: { parts: [...pendingParts] } }] : [],
+              usageMetadata: latestUsageMetadata
+            };
+            pendingParts = [];
+            lastFlushTime = Date.now();
+            hasNewData = false;
+          }
+          if (isDone) break;
         }
         if (Date.now() - lastFlushTime >= 150 && hasNewData) {
           yield {
@@ -15497,33 +15515,50 @@ var init_ai = __esm({
         for (const line of lines) {
           const cleanLine = line.trim();
           if (!cleanLine || !cleanLine.startsWith("data: ")) continue;
-          if (cleanLine === "data: [DONE]") break;
-          try {
-            const json = JSON.parse(cleanLine.substring(6));
-            const delta = json.choices?.[0]?.delta;
-            const usage = json.usage;
-            if (usage) {
-              latestUsageMetadata = {
-                totalTokenCount: usage.total_tokens || (usage.prompt_tokens || 0) + (usage.completion_tokens || 0),
-                promptTokenCount: usage.prompt_tokens || 0,
-                candidatesTokenCount: usage.completion_tokens || 0,
-                cachedContentTokenCount: usage.prompt_tokens_details?.cached_tokens || 0,
-                thoughtsTokenCount: 0
-              };
-              hasNewData = true;
-            }
-            if (delta) {
-              if (delta.thinking) {
-                pendingParts.push({ text: delta.thinking, thought: true });
-                hasNewData = true;
+          let isDone = false;
+          if (cleanLine === "data: [DONE]") {
+            isDone = true;
+          } else {
+            try {
+              const json = JSON.parse(cleanLine.substring(6));
+              const delta = json.choices?.[0]?.delta;
+              const usage = json.usage;
+              if (json.choices?.[0]?.finish_reason) {
+                isDone = true;
               }
-              if (delta.content) {
-                pendingParts.push({ text: delta.content });
+              if (usage) {
+                latestUsageMetadata = {
+                  totalTokenCount: usage.total_tokens || (usage.prompt_tokens || 0) + (usage.completion_tokens || 0),
+                  promptTokenCount: usage.prompt_tokens || 0,
+                  candidatesTokenCount: usage.completion_tokens || 0,
+                  cachedContentTokenCount: usage.prompt_tokens_details?.cached_tokens || 0,
+                  thoughtsTokenCount: 0
+                };
                 hasNewData = true;
               }
+              if (delta) {
+                if (delta.thinking) {
+                  pendingParts.push({ text: delta.thinking, thought: true });
+                  hasNewData = true;
+                }
+                if (delta.content) {
+                  pendingParts.push({ text: delta.content });
+                  hasNewData = true;
+                }
+              }
+            } catch (e) {
             }
-          } catch (e) {
           }
+          if ((isDone || Date.now() - lastFlushTime >= 150) && hasNewData) {
+            yield {
+              candidates: pendingParts.length > 0 ? [{ content: { parts: [...pendingParts] } }] : [],
+              usageMetadata: latestUsageMetadata
+            };
+            pendingParts = [];
+            lastFlushTime = Date.now();
+            hasNewData = false;
+          }
+          if (isDone) break;
         }
         if (Date.now() - lastFlushTime >= 150 && hasNewData) {
           yield {
@@ -15716,9 +15751,14 @@ var init_ai = __esm({
             buffer += decoder.decode(value, { stream: true });
             const lines = buffer.split("\n");
             buffer = lines.pop();
+            let isDone = false;
             for (const line of lines) {
               const trimmed = line.trim();
-              if (!trimmed || trimmed === "data: [DONE]") continue;
+              if (!trimmed) continue;
+              if (trimmed === "data: [DONE]") {
+                isDone = true;
+                break;
+              }
               if (trimmed.startsWith("data: ")) {
                 let json;
                 try {
@@ -15729,6 +15769,9 @@ var init_ai = __esm({
                 if (json.error) {
                   throw new Error(`NVIDIA Stream Error: ${json.error.message || JSON.stringify(json.error)}`);
                 }
+                if (json.choices?.[0]?.finish_reason) {
+                  isDone = true;
+                }
                 try {
                   const usage = json.usage;
                   if (usage) {
@@ -15755,7 +15798,7 @@ var init_ai = __esm({
                 }
               }
             }
-            if (Date.now() - lastFlushTime >= 350 && hasNewData) {
+            if ((isDone || Date.now() - lastFlushTime >= 350) && hasNewData) {
               yield {
                 candidates: pendingParts.length > 0 ? [{ content: { parts: [...pendingParts] } }] : [],
                 usageMetadata: latestUsageMetadata
@@ -15765,6 +15808,7 @@ var init_ai = __esm({
               lastFlushTime = Date.now();
               hasNewData = false;
             }
+            if (isDone) break;
           }
           break;
         } catch (error) {
@@ -15982,43 +16026,51 @@ var init_ai = __esm({
         for (const line of lines) {
           const cleanLine = line.trim();
           if (!cleanLine || !cleanLine.startsWith("data: ")) continue;
-          if (cleanLine === "data: [DONE]") break;
-          try {
-            const json = JSON.parse(cleanLine.substring(6));
-            const delta = json.choices?.[0]?.delta;
-            const usage = json.usage;
-            if (usage) {
-              latestUsageMetadata = {
-                totalTokenCount: usage.total_tokens || (usage.prompt_tokens || 0) + (usage.completion_tokens || 0),
-                promptTokenCount: usage.prompt_tokens || 0,
-                candidatesTokenCount: usage.completion_tokens || 0,
-                cachedContentTokenCount: usage.prompt_tokens_details?.cached_tokens || usage.prompt_tokens_details?.cache_read_input_tokens || usage.cache_read_input_tokens || 0,
-                thoughtsTokenCount: usage.completion_tokens_details?.reasoning_tokens || 0
-              };
-              hasNewData = true;
-            }
-            if (delta) {
-              const thought = delta.reasoning || (delta.reasoning_details ? delta.reasoning_details.map((d) => d.text).join("") : null);
-              if (thought) {
-                pendingParts.push({ text: thought, thought: true });
-                hasNewData = true;
+          let isDone = false;
+          if (cleanLine === "data: [DONE]") {
+            isDone = true;
+          } else {
+            try {
+              const json = JSON.parse(cleanLine.substring(6));
+              const delta = json.choices?.[0]?.delta;
+              const usage = json.usage;
+              if (json.choices?.[0]?.finish_reason) {
+                isDone = true;
               }
-              if (delta.content) {
-                pendingParts.push({ text: delta.content });
+              if (usage) {
+                latestUsageMetadata = {
+                  totalTokenCount: usage.total_tokens || (usage.prompt_tokens || 0) + (usage.completion_tokens || 0),
+                  promptTokenCount: usage.prompt_tokens || 0,
+                  candidatesTokenCount: usage.completion_tokens || 0,
+                  cachedContentTokenCount: usage.prompt_tokens_details?.cached_tokens || usage.prompt_tokens_details?.cache_read_input_tokens || usage.cache_read_input_tokens || 0,
+                  thoughtsTokenCount: usage.completion_tokens_details?.reasoning_tokens || 0
+                };
                 hasNewData = true;
               }
+              if (delta) {
+                const thought = delta.reasoning || (delta.reasoning_details ? delta.reasoning_details.map((d) => d.text).join("") : null);
+                if (thought) {
+                  pendingParts.push({ text: thought, thought: true });
+                  hasNewData = true;
+                }
+                if (delta.content) {
+                  pendingParts.push({ text: delta.content });
+                  hasNewData = true;
+                }
+              }
+            } catch (e) {
             }
-          } catch (e) {
           }
-        }
-        if (Date.now() - lastFlushTime >= 150 && hasNewData) {
-          yield {
-            candidates: pendingParts.length > 0 ? [{ content: { parts: [...pendingParts] } }] : [],
-            usageMetadata: latestUsageMetadata
-          };
-          pendingParts = [];
-          lastFlushTime = Date.now();
-          hasNewData = false;
+          if ((isDone || Date.now() - lastFlushTime >= 150) && hasNewData) {
+            yield {
+              candidates: pendingParts.length > 0 ? [{ content: { parts: [...pendingParts] } }] : [],
+              usageMetadata: latestUsageMetadata
+            };
+            pendingParts = [];
+            lastFlushTime = Date.now();
+            hasNewData = false;
+          }
+          if (isDone) break;
         }
       }
     };
@@ -16091,18 +16143,24 @@ var init_ai = __esm({
           hasNewData = true;
         }
         if (chunk.done) {
-          const evalNs = chunk.prompt_eval_duration || 0;
-          const isCached = evalNs > 0 && evalNs < 75e6;
+          let cachedCount = chunk.prompt_cached_count || 0;
+          if (!cachedCount && chunk.prompt_eval_count) {
+            const evalNs = chunk.prompt_eval_duration || 0;
+            if (evalNs > 0 && evalNs < 75e6) {
+              cachedCount = chunk.prompt_eval_count;
+            }
+          }
           latestUsageMetadata = {
             totalTokenCount: (chunk.prompt_eval_count || 0) + (chunk.eval_count || 0),
             promptTokenCount: chunk.prompt_eval_count || 0,
             candidatesTokenCount: chunk.eval_count || 0,
-            cachedContentTokenCount: chunk.prompt_eval_count && isCached ? chunk.prompt_eval_count : 0,
+            cachedContentTokenCount: cachedCount,
             thoughtsTokenCount: 0
+            // Note: Ollama does not natively return a separate sub-count for thoughts yet
           };
           hasNewData = true;
         }
-        if (Date.now() - lastFlushTime >= 150 && hasNewData) {
+        if (chunk.done || Date.now() - lastFlushTime >= 150 && hasNewData) {
           yield {
             candidates: pendingParts.length > 0 ? [{ content: { parts: [...pendingParts] } }] : [],
             usageMetadata: latestUsageMetadata
@@ -17850,7 +17908,7 @@ ${combinedNudge}`;
                 if (!text || !text.includes("[tool:")) return text;
                 const THINK_OPEN_PH = "___THINK_OPEN_TAG___";
                 const THINK_CLOSE_PH = "___THINK_CLOSE_TAG___";
-                text = text.replaceAll("<think>", THINK_OPEN_PH).replaceAll("</think>", THINK_CLOSE_PH);
+                text = text.replace("<think>", THINK_OPEN_PH).replace("</think>", THINK_CLOSE_PH);
                 text = text.replace(/<(\w+)(?:[^>]*)>\s*([\s\S]*?\[tool:[^\]]*\][\s\S]*?)\s*<\/\1>/gi, (match2, tagName, innerContent) => {
                   if (innerContent && innerContent.includes("[tool:")) return innerContent.trim();
                   return match2;
@@ -19933,6 +19991,7 @@ Error Log can be found in ${path26.join(LOGS_DIR, "agent", "error.log")}`);
         if (lower === "openrouter") return "OpenRouter";
         if (lower === "nvidia") return "NVIDIA";
         if (lower === "mistral") return "Mistral";
+        if (lower === "ollama") return "Ollama";
         return null;
       };
       const envSubagentProvider = normalizeProvider(envSubagentProviderRaw);
@@ -19997,6 +20056,7 @@ TOOL RULES:
 - JSON ESCAPE ALL LITERAL ESCAPE SEQUENCES IN TOOL ARGUMENTS
 - SAME file, MULTIPLE edits? ONE PatchFile (\u226415 blocks) \u2190 PRIORITY
 - Need text or huge files? SearchKeyword > Full Read
+- MUST AVOID UNNECESSARY LARGE-FILE CHUNK READS
 - Restricted Shell Access, No Deletion
 - ONLY valid tools and syntax defined below are allowed
 
@@ -21439,7 +21499,7 @@ function App({ args = [] }) {
         }
         const envModel = process.env.SUBAGENT_MODEL ? process.env.SUBAGENT_MODEL.trim() : null;
         const envProviderRaw = process.env.SUBAGENT_PROVIDER ? process.env.SUBAGENT_PROVIDER.trim() : null;
-        const ALL_PROVIDERS = ["Google", "DeepSeek", "OpenRouter", "NVIDIA", "Mistral"];
+        const ALL_PROVIDERS = ["Google", "DeepSeek", "OpenRouter", "NVIDIA", "Mistral", "Ollama"];
         const normalizeProvider = (pStr) => {
           if (!pStr) return null;
           const lower = pStr.toLowerCase();
@@ -21448,6 +21508,7 @@ function App({ args = [] }) {
           if (lower === "openrouter") return "OpenRouter";
           if (lower === "nvidia") return "NVIDIA";
           if (lower === "mistral") return "Mistral";
+          if (lower === "ollama") return "Ollama";
           return null;
         };
         const envProvider = normalizeProvider(envProviderRaw);

package/model_config.json

@@ -1,5 +1,5 @@
 {
-    "version": 0,
+    "version": 1,
     "release": 20260806,
     "fallbacks": {
         "janitor_default": "gemini-3.1-flash-lite",
@@ -42,6 +42,21 @@
                         "multimodal": false,
                         "desc": "Cost Saving (Text Only) [EXPERIMENTAL]"
                     },
+                    {
+                        "cmd": "ministral-3b-2512",
+                        "multimodal": false,
+                        "desc": "Cost Saving (Text Only) [EXPERIMENTAL]"
+                    },
+                    {
+                        "cmd": "ministral-8b-2512",
+                        "multimodal": false,
+                        "desc": "Cost Saving (Text Only) [EXPERIMENTAL]"
+                    },
+                    {
+                        "cmd": "ministral-14b-2512",
+                        "multimodal": false,
+                        "desc": "Cost Saving (Text Only) [EXPERIMENTAL]"
+                    },
                     {
                         "cmd": "mistral-small-2506",
                         "multimodal": false,
@@ -53,10 +68,15 @@
                         "desc": "Devstral Coding (Text Only) [EXPERIMENTAL]"
                     },
                     {
-                        "cmd": "mistral-medium-2505",
+                        "cmd": "mistral-medium-2508",
                         "multimodal": true,
                         "desc": "Mistral Medium (Multimodal)"
                     },
+                    {
+                        "cmd": "mistral-large-2512",
+                        "multimodal": true,
+                        "desc": "Mistral Large (Multimodal)"
+                    },
                     {
                         "cmd": "labs-leanstral-1-5-1",
                         "multimodal": false,
@@ -73,6 +93,21 @@
                         "multimodal": false,
                         "desc": "Cost Saving (Text Only) [EXPERIMENTAL]"
                     },
+                    {
+                        "cmd": "ministral-3b-2512",
+                        "multimodal": false,
+                        "desc": "Cost Saving (Text Only) [EXPERIMENTAL]"
+                    },
+                    {
+                        "cmd": "ministral-8b-2512",
+                        "multimodal": false,
+                        "desc": "Cost Saving (Text Only) [EXPERIMENTAL]"
+                    },
+                    {
+                        "cmd": "ministral-14b-2512",
+                        "multimodal": false,
+                        "desc": "Cost Saving (Text Only) [EXPERIMENTAL]"
+                    },
                     {
                         "cmd": "mistral-small-2506",
                         "multimodal": false,
@@ -84,10 +119,15 @@
                         "desc": "Devstral Coding (Text Only) [EXPERIMENTAL]"
                     },
                     {
-                        "cmd": "mistral-medium-2505",
+                        "cmd": "mistral-medium-2508",
                         "multimodal": true,
                         "desc": "Mistral Medium (Multimodal)"
                     },
+                    {
+                        "cmd": "mistral-large-2512",
+                        "multimodal": true,
+                        "desc": "Mistral Large (Multimodal)"
+                    },
                     {
                         "cmd": "labs-leanstral-1-5-1",
                         "multimodal": false,

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "fluxflow-cli",
-    "version": "3.19.1",
+    "version": "3.19.2",
     "date": "2026-08-06",
     "description": "A High-Fidelity Agentic CLI with Sub-Agents for the Flux Era.",
     "keywords": [