@stevederico/dotbot 0.31.0 → 0.32.0

package/.claude/settings.local.json

@@ -0,0 +1,7 @@
+{
+  "permissions": {
+    "allow": [
+      "Bash(grep -r \"grok-3\" /Users/sd/Desktop/projects/dotbot --include=\"*.js\" --include=\"*.md\" 2>/dev/null | grep -v node_modules)"
+    ]
+  }
+}

package/CHANGELOG.md

@@ -1,3 +1,9 @@
+0.32
+
+  Rename mlx_local provider to local
+  Rename MLX_LOCAL_URL env var to LOCAL_LLM_URL
+  Remove MLX references from comments and docs
+
 0.31
 
   Document mlx_local provider

package/README.md

@@ -217,7 +217,7 @@ Commands:
   events [--summary]          View audit log
 
 Options:
-  --provider, -p   AI provider: xai, anthropic, openai, ollama, mlx_local (default: xai)
+  --provider, -p   AI provider: xai, anthropic, openai, ollama, local (default: xai)
   --model, -m      Model name (default: grok-4-1-fast-reasoning)
   --system, -s     Custom system prompt (prepended to default)
   --session        Resume a specific session by ID
@@ -236,7 +236,7 @@ Environment Variables:
   ANTHROPIC_API_KEY    API key for Anthropic
   OPENAI_API_KEY       API key for OpenAI
   OLLAMA_BASE_URL      Base URL for Ollama (default: http://localhost:11434)
-  MLX_LOCAL_URL        Base URL for a local MLX-style OpenAI-compatible server (default: http://127.0.0.1:1316/v1)
+  LOCAL_LLM_URL        Base URL for a local OpenAI-compatible LLM server (default: http://127.0.0.1:1316/v1)
 
 Config File:
   ~/.dotbotrc          JSON config for defaults (provider, model, db, sandbox)

package/bin/dotbot.js

@@ -115,7 +115,7 @@ Commands:
   events [--summary]          View audit log
 
 Options:
-  --provider, -p   AI provider: xai, anthropic, openai, ollama, mlx_local (default: xai)
+  --provider, -p   AI provider: xai, anthropic, openai, ollama, local (default: xai)
   --model, -m      Model name (default: grok-4-1-fast-reasoning)
   --system, -s     Custom system prompt (prepended to default)
   --session        Resume a specific session by ID
@@ -134,7 +134,7 @@ Environment Variables:
   ANTHROPIC_API_KEY    API key for Anthropic
   OPENAI_API_KEY       API key for OpenAI
   OLLAMA_BASE_URL      Base URL for Ollama (default: http://localhost:11434)
-  MLX_LOCAL_URL        Base URL for a local MLX-style OpenAI-compatible server (default: http://127.0.0.1:1316/v1)
+  LOCAL_LLM_URL        Base URL for a local local OpenAI-compatible LLM server (default: http://127.0.0.1:1316/v1)
 
 Config File:
   ~/.dotbotrc                 JSON config for defaults (provider, model, db)
@@ -296,7 +296,7 @@ async function getProviderConfig(providerId) {
     return { ...base, apiUrl: `${baseUrl}/api/chat` };
   }
 
-  // Local OpenAI-compatible servers (mlx_local, etc.) don't use API keys —
+  // Local OpenAI-compatible servers (local, etc.) don't use API keys —
   // they're served from localhost and the apiUrl is already baked into the
   // provider config (or overridden via env var inside providers.js).
   if (base.local) {

package/core/agent.js

@@ -162,7 +162,7 @@ export async function* agentLoop({ model, messages, tools, signal, provider, con
       };
 
       // Include tool definitions for non-local providers and local providers
-      // that support native tool calling (e.g., GLM-4.7 via mlx_lm.server v0.30.7+)
+      // that support native tool calling (e.g., GLM-4.7 via local LLM server v0.30.7+)
       if (!targetProvider.local || targetProvider.supportsToolRole) {
         requestBody.tools = toolDefs;
       }
@@ -174,7 +174,7 @@ export async function* agentLoop({ model, messages, tools, signal, provider, con
       };
     };
 
-    // Local providers (ollama, mlx_local): direct fetch, no failover
+    // Local providers (ollama, local): direct fetch, no failover
     if (provider.local) {
       const { url, headers, body } = buildAgentRequest(provider);
       response = await fetch(url, { method: "POST", headers, body, signal });
@@ -210,8 +210,8 @@ export async function* agentLoop({ model, messages, tools, signal, provider, con
       const result = yield* parseAnthropicStream(response, fullContent, toolCalls, signal, activeProvider.id);
       fullContent = result.fullContent;
       toolCalls = result.toolCalls;
-    } else if (activeProvider.id === "mlx_local") {
-      // Local MLX-style OpenAI-compatible server. Models served this way
+    } else if (activeProvider.id === "local") {
+      // Local OpenAI-compatible server. Models served this way
       // may emit output in one of three formats:
       // 1. gpt-oss channel tokens (<|channel|>analysis/final<|message|>)
       // 2. Native reasoning (delta.reasoning from parseOpenAIStream)
@@ -286,7 +286,7 @@ export async function* agentLoop({ model, messages, tools, signal, provider, con
         // the model doesn't use gpt-oss format (e.g. LFM2.5, SmolLM).
         // Flush buffer and switch to passthrough for remaining tokens.
         if (!analysisStarted && !finalMarkerFound && rawBuffer.length > CHANNEL_DETECT_THRESHOLD) {
-          console.log("[mlx_local] no channel tokens after", rawBuffer.length, "chars — switching to passthrough");
+          console.log("[local] no channel tokens after", rawBuffer.length, "chars — switching to passthrough");
           usesPassthrough = true;
           const textEvent = { type: "text_delta", text: rawBuffer };
           validateEvent(textEvent);
@@ -301,7 +301,7 @@ export async function* agentLoop({ model, messages, tools, signal, provider, con
             if (aIdx !== -1) {
               analysisStarted = true;
               lastThinkingYieldPos = aIdx + ANALYSIS_MARKER.length;
-              console.log("[mlx_local] analysis marker found at", aIdx, "| yieldPos:", lastThinkingYieldPos);
+              console.log("[local] analysis marker found at", aIdx, "| yieldPos:", lastThinkingYieldPos);
             }
           }
 
@@ -311,7 +311,7 @@ export async function* agentLoop({ model, messages, tools, signal, provider, con
             if (endIdx !== -1) {
               const chunk = rawBuffer.slice(lastThinkingYieldPos, endIdx);
               if (chunk) {
-                console.log("[mlx_local] thinking (final):", chunk.slice(0, 80));
+                console.log("[local] thinking (final):", chunk.slice(0, 80));
                 const thinkingEvent = {
                   type: "thinking",
                   text: chunk,
@@ -325,7 +325,7 @@ export async function* agentLoop({ model, messages, tools, signal, provider, con
             } else {
               const chunk = rawBuffer.slice(lastThinkingYieldPos);
               if (chunk) {
-                console.log("[mlx_local] thinking (incr):", chunk.slice(0, 80));
+                console.log("[local] thinking (incr):", chunk.slice(0, 80));
                 const thinkingEvent = {
                   type: "thinking",
                   text: chunk,
@@ -341,7 +341,7 @@ export async function* agentLoop({ model, messages, tools, signal, provider, con
           // Check for final channel marker
           const fIdx = rawBuffer.indexOf(FINAL_MARKER);
           if (fIdx !== -1) {
-            console.log("[mlx_local] final marker found at", fIdx, "| bufLen:", rawBuffer.length);
+            console.log("[local] final marker found at", fIdx, "| bufLen:", rawBuffer.length);
             finalMarkerFound = true;
             lastFinalYieldPos = fIdx + FINAL_MARKER.length;
             const pending = rawBuffer.slice(lastFinalYieldPos);
@@ -757,8 +757,8 @@ export async function getOllamaStatus() {
 
 /**
  * Check if a local OpenAI-compatible model server is running and list
- * available models. Defaults to the MLX LM server convention
- * (http://localhost:1316/v1) and can be overridden with MLX_LOCAL_URL.
+ * available models. Defaults to the local LLM server convention
+ * (http://localhost:1316/v1) and can be overridden with LOCAL_LLM_URL.
  *
  * @returns {Promise<{running: boolean, models: Array<{name: string}>}>}
  */
@@ -783,7 +783,7 @@ function stripGptOssTokens(text) {
 }
 
 export async function getMlxLocalStatus() {
-  const baseUrl = (process.env.MLX_LOCAL_URL || 'http://localhost:1316/v1').replace(/\/v1$/, '');
+  const baseUrl = (process.env.LOCAL_LLM_URL || 'http://localhost:1316/v1').replace(/\/v1$/, '');
   try {
     const res = await fetch(`${baseUrl}/v1/models`);
     if (!res.ok) return { running: false, models: [] };

package/core/compaction.js

@@ -12,7 +12,7 @@ const CONTEXT_LIMITS = {
   openai: 120000,
   xai: 120000,
   ollama: 6000,
-  mlx_local: 6000,
+  local: 6000,
 };
 
 /** Number of recent messages to always preserve verbatim. */

package/core/gptoss_tool_parser.js

@@ -12,11 +12,11 @@
  * 3. LFM2.5 native format with markers:
  *    <|tool_call_start|>[tool_name(arg1="value1")]<|tool_call_end|>
  *
- * 4. LFM2.5 bare Pythonic format (markers stripped by mlx_lm.server):
+ * 4. LFM2.5 bare Pythonic format (markers stripped by local LLM server):
  *    [tool_name(arg1="value1", arg2="value2")]
  *
  * Used when the model doesn't support native OpenAI-style tool calling
- * (e.g., mlx_lm.server) and tool definitions are injected via system prompt.
+ * (e.g., local LLM server) and tool definitions are injected via system prompt.
  */
 
 const TOOL_CALL_RE = /<tool_call>([\s\S]*?)<\/tool_call>/g;
@@ -135,7 +135,7 @@ export function parseToolCalls(text) {
     }
   }
 
-  // Format 4: [func_name(key="val")]  (bare Pythonic, markers stripped by mlx_lm.server)
+  // Format 4: [func_name(key="val")]  (bare Pythonic, markers stripped by local LLM server)
   if (calls.length === 0) {
     BARE_PYTHONIC_RE.lastIndex = 0;
     while ((match = BARE_PYTHONIC_RE.exec(text)) !== null) {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@stevederico/dotbot",
-  "version": "0.31.0",
+  "version": "0.32.0",
   "description": "AI agent CLI and library for Node.js — streaming, multi-provider, tool execution, autonomous tasks",
   "type": "module",
   "main": "index.js",

package/test/agent.test.js

@@ -3,7 +3,7 @@ import assert from 'node:assert';
 import { agentLoop } from '../core/agent.js';
 
 /**
- * Regression tests for the mlx_local provider branch of agentLoop.
+ * Regression tests for the local provider branch of agentLoop.
  *
  * These cover the flush branch added in 0.30 that handles short plain-text
  * responses from local models that never emit gpt-oss channel tokens
@@ -13,13 +13,13 @@ import { agentLoop } from '../core/agent.js';
  */
 
 /**
- * Build a minimal mlx_local-style provider for agentLoop tests.
- * The `id` must be "mlx_local" to hit the buffered-parsing branch,
+ * Build a minimal local-style provider for agentLoop tests.
+ * The `id` must be "local" to hit the buffered-parsing branch,
  * and `local: true` skips the failover path for a direct fetch.
  */
 function makeLocalProvider() {
   return {
-    id: 'mlx_local',
+    id: 'local',
     name: 'Test Local',
     apiUrl: 'http://127.0.0.1:1316/v1',
     endpoint: '/chat/completions',
@@ -61,7 +61,7 @@ function stubFetch(response) {
   return () => { globalThis.fetch = original; };
 }
 
-describe('agentLoop — mlx_local short plain-text response flush', () => {
+describe('agentLoop — local short plain-text response flush', () => {
   let restoreFetch;
 
   afterEach(() => {

package/utils/providers.js

@@ -133,14 +133,14 @@ export const AI_PROVIDERS = {
     }),
     formatResponse: (data) => data.choices?.[0]?.message?.content
   },
-  mlx_local: {
-    // Local MLX-style OpenAI-compatible server (e.g. mlx_lm.server, LM Studio,
-    // vLLM, llama.cpp server). Routes through the `mlx_local` branch in
+  local: {
+    // Local OpenAI-compatible server (e.g. local LLM server, LM Studio,
+    // vLLM, llama.cpp server). Routes through the `local` branch in
     // core/agent.js which auto-detects gpt-oss channel tokens, native
-    // reasoning, and plain-text responses. Override the URL with MLX_LOCAL_URL.
-    id: 'mlx_local',
-    name: 'Local (MLX)',
-    apiUrl: process.env.MLX_LOCAL_URL || 'http://127.0.0.1:1316/v1',
+    // reasoning, and plain-text responses. Override the URL with LOCAL_LLM_URL.
+    id: 'local',
+    name: 'Local',
+    apiUrl: process.env.LOCAL_LLM_URL || 'http://127.0.0.1:1316/v1',
     defaultModel: '',
     models: [],
     local: true,