codemaxxing 0.1.10 → 0.1.12

package/dist/index.js

@@ -29,6 +29,7 @@ function formatTimeAgo(date) {
 // ── Slash Commands ──
 const SLASH_COMMANDS = [
     { cmd: "/help", desc: "show commands" },
+    { cmd: "/connect", desc: "retry LLM connection" },
     { cmd: "/login", desc: "set up authentication" },
     { cmd: "/map", desc: "show repository map" },
     { cmd: "/reset", desc: "clear conversation" },
@@ -127,115 +128,124 @@ function App() {
         pasteEvents.on("paste", handler);
         return () => { pasteEvents.off("paste", handler); };
     }, []);
-    // Initialize agent
-    useEffect(() => {
-        (async () => {
-            const cliArgs = parseCLIArgs();
-            const rawConfig = loadConfig();
-            const config = applyOverrides(rawConfig, cliArgs);
-            let provider = config.provider;
-            const info = [];
-            if (provider.model === "auto" || (provider.baseUrl === "http://localhost:1234/v1" && !cliArgs.baseUrl)) {
-                info.push("Detecting local LLM server...");
+    // Connect/reconnect to LLM provider
+    const connectToProvider = useCallback(async (isRetry = false) => {
+        const cliArgs = parseCLIArgs();
+        const rawConfig = loadConfig();
+        const config = applyOverrides(rawConfig, cliArgs);
+        let provider = config.provider;
+        const info = [];
+        if (isRetry) {
+            info.push("Retrying connection...");
+            setConnectionInfo([...info]);
+        }
+        if (provider.model === "auto" || (provider.baseUrl === "http://localhost:1234/v1" && !cliArgs.baseUrl)) {
+            info.push("Detecting local LLM server...");
+            setConnectionInfo([...info]);
+            const detected = await detectLocalProvider();
+            if (detected) {
+                // Keep CLI model override if specified
+                if (cliArgs.model)
+                    detected.model = cliArgs.model;
+                provider = detected;
+                info.push(`✔ Connected to ${provider.baseUrl} → ${provider.model}`);
                 setConnectionInfo([...info]);
-                const detected = await detectLocalProvider();
-                if (detected) {
-                    // Keep CLI model override if specified
-                    if (cliArgs.model)
-                        detected.model = cliArgs.model;
-                    provider = detected;
-                    info.push(`✔ Connected to ${provider.baseUrl} → ${provider.model}`);
-                    setConnectionInfo([...info]);
-                }
-                else {
-                    info.push("✗ No local LLM server found. Start LM Studio or Ollama.");
-                    info.push("  Use --base-url and --api-key to connect to a remote provider.");
-                    info.push("  Type /login to authenticate with a cloud provider.");
-                    setConnectionInfo([...info]);
-                    setReady(true);
-                    return;
-                }
             }
             else {
-                info.push(`Provider: ${provider.baseUrl}`);
-                info.push(`Model: ${provider.model}`);
-                setConnectionInfo([...info]);
-            }
-            const cwd = process.cwd();
-            // Git info
-            if (isGitRepo(cwd)) {
-                const branch = getBranch(cwd);
-                const status = getStatus(cwd);
-                info.push(`Git: ${branch} (${status})`);
+                info.push("✗ No local LLM server found.");
+                info.push("  /connect  — retry after starting LM Studio or Ollama");
+                info.push("  /login    — authenticate with a cloud provider");
                 setConnectionInfo([...info]);
-            }
-            const a = new CodingAgent({
-                provider,
-                cwd,
-                maxTokens: config.defaults.maxTokens,
-                autoApprove: config.defaults.autoApprove,
-                onToken: (token) => {
-                    // Switch from big spinner to streaming mode
-                    setLoading(false);
-                    setStreaming(true);
-                    // Update the current streaming response in-place
-                    setMessages((prev) => {
-                        const lastIdx = prev.length - 1;
-                        const last = prev[lastIdx];
-                        if (last && last.type === "response" && last._streaming) {
-                            return [
-                                ...prev.slice(0, lastIdx),
-                                { ...last, text: last.text + token },
-                            ];
-                        }
-                        // First token of a new response
-                        return [...prev, { id: msgId++, type: "response", text: token, _streaming: true }];
-                    });
-                },
-                onToolCall: (name, args) => {
-                    setLoading(true);
-                    setSpinnerMsg("Executing tools...");
-                    const argStr = Object.entries(args)
-                        .map(([k, v]) => {
-                        const val = String(v);
-                        return val.length > 60 ? val.slice(0, 60) + "..." : val;
-                    })
-                        .join(", ");
-                    addMsg("tool", `${name}(${argStr})`);
-                },
-                onToolResult: (_name, result) => {
-                    const numLines = result.split("\n").length;
-                    const size = result.length > 1024 ? `${(result.length / 1024).toFixed(1)}KB` : `${result.length}B`;
-                    addMsg("tool-result", `└ ${numLines} lines (${size})`);
-                },
-                onThinking: (text) => {
-                    if (text.length > 0) {
-                        addMsg("info", `💭 Thought for ${text.split(/\s+/).length} words`);
+                setReady(true);
+                return;
+            }
+        }
+        else {
+            info.push(`Provider: ${provider.baseUrl}`);
+            info.push(`Model: ${provider.model}`);
+            setConnectionInfo([...info]);
+        }
+        const cwd = process.cwd();
+        // Git info
+        if (isGitRepo(cwd)) {
+            const branch = getBranch(cwd);
+            const status = getStatus(cwd);
+            info.push(`Git: ${branch} (${status})`);
+            setConnectionInfo([...info]);
+        }
+        const a = new CodingAgent({
+            provider,
+            cwd,
+            maxTokens: config.defaults.maxTokens,
+            autoApprove: config.defaults.autoApprove,
+            onToken: (token) => {
+                // Switch from big spinner to streaming mode
+                setLoading(false);
+                setStreaming(true);
+                // Update the current streaming response in-place
+                setMessages((prev) => {
+                    const lastIdx = prev.length - 1;
+                    const last = prev[lastIdx];
+                    if (last && last.type === "response" && last._streaming) {
+                        return [
+                            ...prev.slice(0, lastIdx),
+                            { ...last, text: last.text + token },
+                        ];
                     }
-                },
-                onGitCommit: (message) => {
-                    addMsg("info", `📝 Auto-committed: ${message}`);
-                },
-                onContextCompressed: (oldTokens, newTokens) => {
-                    const saved = oldTokens - newTokens;
-                    const savedStr = saved >= 1000 ? `${(saved / 1000).toFixed(1)}k` : String(saved);
-                    addMsg("info", `📦 Context compressed (~${savedStr} tokens freed)`);
-                },
-                contextCompressionThreshold: config.defaults.contextCompressionThreshold,
-                onToolApproval: (name, args, diff) => {
-                    return new Promise((resolve) => {
-                        setApproval({ tool: name, args, diff, resolve });
-                        setLoading(false);
-                    });
-                },
-            });
-            // Initialize async context (repo map)
-            await a.init();
-            setAgent(a);
-            setModelName(provider.model);
-            providerRef.current = { baseUrl: provider.baseUrl, apiKey: provider.apiKey };
-            setReady(true);
-        })();
+                    // First token of a new response
+                    return [...prev, { id: msgId++, type: "response", text: token, _streaming: true }];
+                });
+            },
+            onToolCall: (name, args) => {
+                setLoading(true);
+                setSpinnerMsg("Executing tools...");
+                const argStr = Object.entries(args)
+                    .map(([k, v]) => {
+                    const val = String(v);
+                    return val.length > 60 ? val.slice(0, 60) + "..." : val;
+                })
+                    .join(", ");
+                addMsg("tool", `${name}(${argStr})`);
+            },
+            onToolResult: (_name, result) => {
+                const numLines = result.split("\n").length;
+                const size = result.length > 1024 ? `${(result.length / 1024).toFixed(1)}KB` : `${result.length}B`;
+                addMsg("tool-result", `└ ${numLines} lines (${size})`);
+            },
+            onThinking: (text) => {
+                if (text.length > 0) {
+                    addMsg("info", `💭 Thought for ${text.split(/\s+/).length} words`);
+                }
+            },
+            onGitCommit: (message) => {
+                addMsg("info", `📝 Auto-committed: ${message}`);
+            },
+            onContextCompressed: (oldTokens, newTokens) => {
+                const saved = oldTokens - newTokens;
+                const savedStr = saved >= 1000 ? `${(saved / 1000).toFixed(1)}k` : String(saved);
+                addMsg("info", `📦 Context compressed (~${savedStr} tokens freed)`);
+            },
+            contextCompressionThreshold: config.defaults.contextCompressionThreshold,
+            onToolApproval: (name, args, diff) => {
+                return new Promise((resolve) => {
+                    setApproval({ tool: name, args, diff, resolve });
+                    setLoading(false);
+                });
+            },
+        });
+        // Initialize async context (repo map)
+        await a.init();
+        setAgent(a);
+        setModelName(provider.model);
+        providerRef.current = { baseUrl: provider.baseUrl, apiKey: provider.apiKey };
+        setReady(true);
+        if (isRetry) {
+            addMsg("info", `✅ Connected to ${provider.model}`);
+        }
+    }, []);
+    // Initialize agent on mount
+    useEffect(() => {
+        connectToProvider(false);
     }, []);
     function addMsg(type, text) {
         setMessages((prev) => [...prev, { id: msgId++, type, text }]);
@@ -298,10 +308,16 @@ function App() {
             setLoginPickerIndex(0);
             return;
         }
+        if (trimmed === "/connect") {
+            addMsg("info", "🔄 Reconnecting...");
+            await connectToProvider(true);
+            return;
+        }
         if (trimmed === "/help") {
             addMsg("info", [
                 "Commands:",
                 "  /help      — show this",
+                "  /connect   — retry LLM connection",
                 "  /login     — authentication setup (run codemaxxing login in terminal)",
                 "  /model     — switch model mid-session",
                 "  /models    — list available models",

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "codemaxxing",
-  "version": "0.1.10",
+  "version": "0.1.12",
   "description": "Open-source terminal coding agent. Connect any LLM. Max your code.",
   "main": "dist/index.js",
   "bin": {

package/src/index.tsx

@@ -29,6 +29,7 @@ function formatTimeAgo(date: Date): string {
 // ── Slash Commands ──
 const SLASH_COMMANDS = [
   { cmd: "/help", desc: "show commands" },
+  { cmd: "/connect", desc: "retry LLM connection" },
   { cmd: "/login", desc: "set up authentication" },
   { cmd: "/map", desc: "show repository map" },
   { cmd: "/reset", desc: "clear conversation" },
@@ -163,121 +164,132 @@ function App() {
     return () => { pasteEvents.off("paste", handler); };
   }, []);
 
-  // Initialize agent
-  useEffect(() => {
-    (async () => {
-      const cliArgs = parseCLIArgs();
-      const rawConfig = loadConfig();
-      const config = applyOverrides(rawConfig, cliArgs);
-      let provider = config.provider;
-      const info: string[] = [];
-
-      if (provider.model === "auto" || (provider.baseUrl === "http://localhost:1234/v1" && !cliArgs.baseUrl)) {
-        info.push("Detecting local LLM server...");
+  // Connect/reconnect to LLM provider
+  const connectToProvider = useCallback(async (isRetry = false) => {
+    const cliArgs = parseCLIArgs();
+    const rawConfig = loadConfig();
+    const config = applyOverrides(rawConfig, cliArgs);
+    let provider = config.provider;
+    const info: string[] = [];
+
+    if (isRetry) {
+      info.push("Retrying connection...");
+      setConnectionInfo([...info]);
+    }
+
+    if (provider.model === "auto" || (provider.baseUrl === "http://localhost:1234/v1" && !cliArgs.baseUrl)) {
+      info.push("Detecting local LLM server...");
+      setConnectionInfo([...info]);
+      const detected = await detectLocalProvider();
+      if (detected) {
+        // Keep CLI model override if specified
+        if (cliArgs.model) detected.model = cliArgs.model;
+        provider = detected;
+        info.push(`✔ Connected to ${provider.baseUrl} → ${provider.model}`);
         setConnectionInfo([...info]);
-        const detected = await detectLocalProvider();
-        if (detected) {
-          // Keep CLI model override if specified
-          if (cliArgs.model) detected.model = cliArgs.model;
-          provider = detected;
-          info.push(`✔ Connected to ${provider.baseUrl} → ${provider.model}`);
-          setConnectionInfo([...info]);
-        } else {
-          info.push("✗ No local LLM server found. Start LM Studio or Ollama.");
-          info.push("  Use --base-url and --api-key to connect to a remote provider.");
-          info.push("  Type /login to authenticate with a cloud provider.");
-          setConnectionInfo([...info]);
-          setReady(true);
-          return;
-        }
       } else {
-        info.push(`Provider: ${provider.baseUrl}`);
-        info.push(`Model: ${provider.model}`);
+        info.push("✗ No local LLM server found.");
+        info.push("  /connect  — retry after starting LM Studio or Ollama");
+        info.push("  /login    — authenticate with a cloud provider");
         setConnectionInfo([...info]);
+        setReady(true);
+        return;
       }
+    } else {
+      info.push(`Provider: ${provider.baseUrl}`);
+      info.push(`Model: ${provider.model}`);
+      setConnectionInfo([...info]);
+    }
 
-      const cwd = process.cwd();
+    const cwd = process.cwd();
 
-      // Git info
-      if (isGitRepo(cwd)) {
-        const branch = getBranch(cwd);
-        const status = getStatus(cwd);
-        info.push(`Git: ${branch} (${status})`);
-        setConnectionInfo([...info]);
-      }
+    // Git info
+    if (isGitRepo(cwd)) {
+      const branch = getBranch(cwd);
+      const status = getStatus(cwd);
+      info.push(`Git: ${branch} (${status})`);
+      setConnectionInfo([...info]);
+    }
 
-      const a = new CodingAgent({
-        provider,
-        cwd,
-        maxTokens: config.defaults.maxTokens,
-        autoApprove: config.defaults.autoApprove,
-        onToken: (token) => {
-          // Switch from big spinner to streaming mode
-          setLoading(false);
-          setStreaming(true);
-
-          // Update the current streaming response in-place
-          setMessages((prev) => {
-            const lastIdx = prev.length - 1;
-            const last = prev[lastIdx];
-
-            if (last && last.type === "response" && (last as any)._streaming) {
-              return [
-                ...prev.slice(0, lastIdx),
-                { ...last, text: last.text + token },
-              ];
-            }
-
-            // First token of a new response
-            return [...prev, { id: msgId++, type: "response" as const, text: token, _streaming: true } as any];
-          });
-        },
-        onToolCall: (name, args) => {
-          setLoading(true);
-          setSpinnerMsg("Executing tools...");
-          const argStr = Object.entries(args)
-            .map(([k, v]) => {
-              const val = String(v);
-              return val.length > 60 ? val.slice(0, 60) + "..." : val;
-            })
-            .join(", ");
-          addMsg("tool", `${name}(${argStr})`);
-        },
-        onToolResult: (_name, result) => {
-          const numLines = result.split("\n").length;
-          const size = result.length > 1024 ? `${(result.length / 1024).toFixed(1)}KB` : `${result.length}B`;
-          addMsg("tool-result", `└ ${numLines} lines (${size})`);
-        },
-        onThinking: (text) => {
-          if (text.length > 0) {
-            addMsg("info", `💭 Thought for ${text.split(/\s+/).length} words`);
+    const a = new CodingAgent({
+      provider,
+      cwd,
+      maxTokens: config.defaults.maxTokens,
+      autoApprove: config.defaults.autoApprove,
+      onToken: (token) => {
+        // Switch from big spinner to streaming mode
+        setLoading(false);
+        setStreaming(true);
+
+        // Update the current streaming response in-place
+        setMessages((prev) => {
+          const lastIdx = prev.length - 1;
+          const last = prev[lastIdx];
+
+          if (last && last.type === "response" && (last as any)._streaming) {
+            return [
+              ...prev.slice(0, lastIdx),
+              { ...last, text: last.text + token },
+            ];
           }
-        },
-        onGitCommit: (message) => {
-          addMsg("info", `📝 Auto-committed: ${message}`);
-        },
-        onContextCompressed: (oldTokens, newTokens) => {
-          const saved = oldTokens - newTokens;
-          const savedStr = saved >= 1000 ? `${(saved / 1000).toFixed(1)}k` : String(saved);
-          addMsg("info", `📦 Context compressed (~${savedStr} tokens freed)`);
-        },
-        contextCompressionThreshold: config.defaults.contextCompressionThreshold,
-        onToolApproval: (name, args, diff) => {
-          return new Promise((resolve) => {
-            setApproval({ tool: name, args, diff, resolve });
-            setLoading(false);
-          });
-        },
-      });
 
-      // Initialize async context (repo map)
-      await a.init();
+          // First token of a new response
+          return [...prev, { id: msgId++, type: "response" as const, text: token, _streaming: true } as any];
+        });
+      },
+      onToolCall: (name, args) => {
+        setLoading(true);
+        setSpinnerMsg("Executing tools...");
+        const argStr = Object.entries(args)
+          .map(([k, v]) => {
+            const val = String(v);
+            return val.length > 60 ? val.slice(0, 60) + "..." : val;
+          })
+          .join(", ");
+        addMsg("tool", `${name}(${argStr})`);
+      },
+      onToolResult: (_name, result) => {
+        const numLines = result.split("\n").length;
+        const size = result.length > 1024 ? `${(result.length / 1024).toFixed(1)}KB` : `${result.length}B`;
+        addMsg("tool-result", `└ ${numLines} lines (${size})`);
+      },
+      onThinking: (text) => {
+        if (text.length > 0) {
+          addMsg("info", `💭 Thought for ${text.split(/\s+/).length} words`);
+        }
+      },
+      onGitCommit: (message) => {
+        addMsg("info", `📝 Auto-committed: ${message}`);
+      },
+      onContextCompressed: (oldTokens, newTokens) => {
+        const saved = oldTokens - newTokens;
+        const savedStr = saved >= 1000 ? `${(saved / 1000).toFixed(1)}k` : String(saved);
+        addMsg("info", `📦 Context compressed (~${savedStr} tokens freed)`);
+      },
+      contextCompressionThreshold: config.defaults.contextCompressionThreshold,
+      onToolApproval: (name, args, diff) => {
+        return new Promise((resolve) => {
+          setApproval({ tool: name, args, diff, resolve });
+          setLoading(false);
+        });
+      },
+    });
+
+    // Initialize async context (repo map)
+    await a.init();
+
+    setAgent(a);
+    setModelName(provider.model);
+    providerRef.current = { baseUrl: provider.baseUrl, apiKey: provider.apiKey };
+    setReady(true);
+    if (isRetry) {
+      addMsg("info", `✅ Connected to ${provider.model}`);
+    }
+  }, []);
 
-      setAgent(a);
-      setModelName(provider.model);
-      providerRef.current = { baseUrl: provider.baseUrl, apiKey: provider.apiKey };
-      setReady(true);
-    })();
+  // Initialize agent on mount
+  useEffect(() => {
+    connectToProvider(false);
   }, []);
 
   function addMsg(type: ChatMessage["type"], text: string) {
@@ -347,10 +359,16 @@ function App() {
       setLoginPickerIndex(0);
       return;
     }
+    if (trimmed === "/connect") {
+      addMsg("info", "🔄 Reconnecting...");
+      await connectToProvider(true);
+      return;
+    }
     if (trimmed === "/help") {
       addMsg("info", [
         "Commands:",
         "  /help      — show this",
+        "  /connect   — retry LLM connection",
         "  /login     — authentication setup (run codemaxxing login in terminal)",
         "  /model     — switch model mid-session",
         "  /models    — list available models",