miii-agent 0.1.29 → 0.1.31

package/README.md

@@ -1,9 +1,9 @@
-<h1 align="center">miii</h1>
+<h1 align="center">miii — Local AI Coding Agent for Your Terminal</h1>
 
 <p align="center">
-  <strong>Cursor / Claude Code, but local.</strong><br>
-  An offline AI pair-programmer in your terminal, powered by Ollama.<br>
-  Private by default. Free forever.
+  <strong>An open-source, offline alternative to Claude Code, Cursor, and GitHub Copilot.</strong><br>
+  A private AI pair-programmer in your terminal, powered by Ollama and any local LLM.<br>
+  Private by default. Free forever. Works offline.
 </p>
 
 <p align="center">
@@ -15,26 +15,59 @@
 </p>
 
 <p align="center">
-  <img src="demo3.gif" alt="miii demo">
+  <img src="demo3.gif" alt="miii local AI coding agent terminal demo powered by Ollama">
+</p>
+
+<p align="center">
+  🔒 <strong>100% local</strong> — your code never leaves your machine &nbsp;·&nbsp;
+  💸 <strong>Free</strong> — no API keys, no per-token billing &nbsp;·&nbsp;
+  ⚡ <strong>Offline</strong> — runs on your own GPU
 </p>
 
 ---
 
-## What is this?
+## What is miii? — a local AI coding agent
 
 miii lives in your terminal and codes alongside you — reading files, writing features, running tests, fixing bugs. The twist: it runs on **your** hardware, powered by [Ollama](https://ollama.com) (or any local OpenAI-compatible server like [llama.cpp](https://github.com/ggml-org/llama.cpp) / [LM Studio](https://lmstudio.ai)).
 
-Your code never leaves your disk. There's nothing to log in to. Pull a model, type `miii`, go.
+Your code never leaves your disk. There's nothing to log in to. Pull a model, type `miii`, go. It's the open-source, offline answer to cloud coding assistants like Claude Code, Cursor, and GitHub Copilot.
 
-## Try it in 30 seconds
+## Install (macOS, Linux, Windows)
+
+**macOS / Linux:**
 
 ```bash
 ollama pull qwen2.5-coder:14b   # any coding model works
+```
+
+**Which model should I use?**
+- **Low VRAM (8GB):** `qwen2.5-coder:7b` (Fast, capable)
+- **Mid VRAM (16-24GB):** `qwen2.5-coder:14b` (Sweet spot)
+- **High VRAM (48GB+):** `qwen2.5-coder:32b` (Powerhouse)
+
+```bash
 curl -fsSL https://raw.githubusercontent.com/maruakshay/miii-cli/main/install.sh | sh
 miii
 ```
+*(The installer downloads the pre-compiled binary and adds it to your local path)*
+
+**Windows (PowerShell):**
+
+```powershell
+ollama pull qwen2.5-coder:14b
+irm https://raw.githubusercontent.com/maruakshay/miii-cli/main/install.ps1 | iex
+miii
+```
+
+Prefer npm? `npm install -g miii-agent` works on every platform.
 
-Prefer npm? `npm install -g miii-agent`.
+> **Install failing on permissions?** Your global npm prefix isn't writable. The
+> installer retries with `sudo` where available; otherwise point npm at a
+> user-owned prefix and re-run:
+> ```bash
+> npm config set prefix "$HOME/.npm-global"
+> export PATH="$HOME/.npm-global/bin:$PATH"   # add to ~/.bashrc or ~/.zshrc
+> ```
 
 Then just talk to it:
 
@@ -59,7 +92,7 @@ miii --version  # what you're running
 Opt out of background updates by adding `"autoUpdate": false` to `~/.miii/config.json`,
 or re-run the install script (`curl … | sh`) any time to update by hand.
 
-## Why local-first?
+## Why local-first? Private, free, offline
 
 Most "AI coding tools" are just wrappers around a cloud API — slow, metered, and they ship your private codebase to someone else's server.
 
@@ -71,13 +104,16 @@ Most "AI coding tools" are just wrappers around a cloud API — slow, metered, a
 | Offline      | No                    | Yes                          |
 | Latency      | Network + queue       | Your GPU only                |
 
-It doesn't just chat, either — it decomposes the problem, calls tools, and checks its own work before claiming victory.
+It doesn't just chat, either — it follows a **Plan $\rightarrow$ Act $\rightarrow$ Observe** loop:
+1. **Plan**: Decomposes the problem into a sequence of concrete steps.
+2. **Act**: Calls the necessary tools to gather context or modify code.
+3. **Observe**: Verifies the result and adjusts the plan until the goal is met.
 
 ## Five letters, five ideas
 
 **s**mall · **s**imple · **s**mart · **s**trategic · **s**emantic — a tiny codebase you can read in an afternoon, no config ceremony, plans before it acts, and operates on the *meaning* of your code, not blind text matching.
 
-## A few things that make it fun
+## Features
 
 - **🧪 `miii doctor`** — not every local model can drive an agent. Doctor runs your models through real engineering tasks and tells you which ones actually deliver.
   ```bash
@@ -171,6 +207,15 @@ The model pages through the middle with ranged `read_file` reads. Spill files ar
 <details>
 <summary><strong>Development</strong></summary>
 
+**Project Architecture:**
+```text
+src/
+ ├── agent/    # The core reasoning loop
+ ├── tools/    # Implementation of read/write/bash
+ ├── terminal/ # UI and input handling
+ └── config/   # Settings and provider logic
+```
+
 ```bash
 git clone https://github.com/maruakshay/miii-cli.git
 cd miii-cli
@@ -193,6 +238,26 @@ npm run build && npm link   # restore later with: npm install -g miii-agent
 
 ---
 
+## FAQ
+
+**Does miii work without internet?**
+Yes. Once you've pulled a model with Ollama, miii runs fully offline. No network calls, no account, no cloud.
+
+**Is my code sent anywhere?**
+No. Every file read, edit, and model inference happens on your machine. Your codebase never leaves your disk.
+
+**Which model is best for coding?**
+Depends on VRAM: `qwen2.5-coder:7b` (8GB), `qwen2.5-coder:14b` (16–24GB, the sweet spot), `qwen2.5-coder:32b` (48GB+). Run `miii doctor` to grade your installed models on real engineering tasks.
+
+**How is miii different from Claude Code, Cursor, or GitHub Copilot?**
+Claude Code, Cursor, and Copilot are cloud services — metered, account-gated, and they ship your code to a third-party server. miii is open-source, free, and runs entirely on your hardware. Same terminal-agent workflow as Claude Code, but on your own local model.
+
+**How is it different from Continue.dev?**
+Continue.dev is an IDE extension. miii is a standalone terminal agent — no editor required — with a Plan → Act → Observe loop, permission-gated tools, and lossless output spill built in.
+
+**Do I need a GPU?**
+No, but it helps. Smaller models run on CPU; a GPU makes larger models fast enough for real work.
+
 ## Status
 
 **MVP.** Core agent loop is stable; actively refining tool execution, streaming, and the permission model. PRs welcome — fork it, break it, improve it.

package/dist/cli.js

@@ -36,7 +36,8 @@ function migrate(raw) {
     effort: raw.effort,
     providers,
     modelContexts: raw.modelContexts,
-    autoUpdate: raw.autoUpdate
+    autoUpdate: raw.autoUpdate,
+    numCtxCap: raw.numCtxCap
   };
 }
 function autoUpdateEnabled(cfg = loadConfig()) {
@@ -50,6 +51,17 @@ function readRawConfig() {
     return {};
   }
 }
+function configError() {
+  if (!existsSync(CONFIG_PATH)) return null;
+  try {
+    JSON.parse(readFileSync(CONFIG_PATH, "utf-8"));
+    return null;
+  } catch (err) {
+    const msg = err instanceof Error ? err.message : String(err);
+    return `miii: ignoring unreadable ${CONFIG_PATH} (${msg}).
+      Running with defaults. Fix the JSON or delete the file to reset.`;
+  }
+}
 function loadConfig() {
   return migrate(readRawConfig());
 }
@@ -88,10 +100,11 @@ function setModelContexts(contexts) {
   const raw = readRawConfig();
   saveConfig({ ...raw, modelContexts: { ...raw.modelContexts, ...contexts } });
 }
-var EFFORT_OPTIONS, CONFIG_DIR, CONFIG_PATH;
+var DEFAULT_NUM_CTX_CAP, EFFORT_OPTIONS, CONFIG_DIR, CONFIG_PATH;
 var init_config = __esm({
   "src/config.ts"() {
     "use strict";
+    DEFAULT_NUM_CTX_CAP = 16384;
     EFFORT_OPTIONS = {
       low: { temperature: 0.2, num_predict: 8192 },
       medium: { temperature: 0.7, num_predict: 16384 },
@@ -535,9 +548,130 @@ var init_client = __esm({
   }
 });
 
-// src/tools/paths.ts
-import { resolve, relative as relative2, isAbsolute, sep, join as join4 } from "path";
+// src/permissions/policy.ts
+import { readFileSync as readFileSync3, writeFileSync as writeFileSync3, mkdirSync as mkdirSync3, existsSync as existsSync3, renameSync } from "fs";
+import { join as join4 } from "path";
 import { homedir as homedir3 } from "os";
+function loadRules() {
+  if (!existsSync3(RULES_PATH)) return [];
+  try {
+    const data = JSON.parse(readFileSync3(RULES_PATH, "utf-8"));
+    return Array.isArray(data.rules) ? data.rules : [];
+  } catch {
+    return [];
+  }
+}
+function saveRules(rules) {
+  mkdirSync3(RULES_DIR, { recursive: true });
+  const tmp = RULES_PATH + ".tmp";
+  writeFileSync3(tmp, JSON.stringify({ rules }, null, 2), "utf-8");
+  renameSync(tmp, RULES_PATH);
+}
+function addRule(tool, pattern) {
+  const rules = loadRules();
+  if (rules.some((r) => r.tool === tool && r.pattern === pattern)) return;
+  rules.push({ tool, pattern });
+  saveRules(rules);
+}
+function subjectFor(toolName, input) {
+  const obj = input ?? {};
+  if (toolName === "run_bash") return typeof obj.command === "string" ? obj.command : "";
+  if (typeof obj.path === "string") return obj.path;
+  return "";
+}
+function generalizeCommand(command) {
+  const trimmed = command.trim();
+  const tokens = trimmed.split(/\s+/);
+  if (tokens.length === 0 || tokens[0] === "") return command;
+  const prog = tokens[0];
+  if (NEVER_GENERALIZE.has(prog)) return trimmed;
+  if (prog === "git" && tokens.length > 1 && DESTRUCTIVE_GIT_SUBCOMMANDS.has(tokens[1])) {
+    return trimmed;
+  }
+  const prefixLen = WRAPPER_PROGRAMS.has(prog) && tokens.length > 1 ? 2 : 1;
+  const prefix = tokens.slice(0, prefixLen).join(" ");
+  return `${prefix} *`;
+}
+function patternToPersist(toolName, subject) {
+  return toolName === "run_bash" ? generalizeCommand(subject) : subject;
+}
+function globToRegExp(glob2) {
+  const escaped = glob2.replace(/[.+^${}()|[\]\\]/g, "\\$&");
+  const pattern = escaped.replace(/\*/g, ".*").replace(/\?/g, ".");
+  return new RegExp(`^${pattern}$`);
+}
+function matches(rule, toolName, subject) {
+  if (rule.tool !== toolName) return false;
+  try {
+    return globToRegExp(rule.pattern).test(subject);
+  } catch {
+    return false;
+  }
+}
+async function check(toolName, input, ctx) {
+  if (ALWAYS_ALLOW.has(toolName)) return "allow";
+  const subject = subjectFor(toolName, input);
+  const rules = loadRules();
+  if (rules.some((r) => matches(r, toolName, subject))) return "allow";
+  const answer = await ctx.ask(toolName, input);
+  if (answer === "no") return "deny";
+  if (answer === "always") addRule(toolName, patternToPersist(toolName, subject));
+  return "allow";
+}
+var RULES_DIR, RULES_PATH, WRAPPER_PROGRAMS, NEVER_GENERALIZE, DESTRUCTIVE_GIT_SUBCOMMANDS, ALWAYS_ALLOW;
+var init_policy = __esm({
+  "src/permissions/policy.ts"() {
+    "use strict";
+    RULES_DIR = join4(homedir3(), ".miii");
+    RULES_PATH = join4(RULES_DIR, "permissions.json");
+    WRAPPER_PROGRAMS = /* @__PURE__ */ new Set([
+      "npm",
+      "npx",
+      "pnpm",
+      "yarn",
+      "brew",
+      "pip",
+      "pip3",
+      "cargo",
+      "docker",
+      "kubectl",
+      "go",
+      "git"
+    ]);
+    NEVER_GENERALIZE = /* @__PURE__ */ new Set([
+      "rm",
+      "rmdir",
+      "dd",
+      "mkfs",
+      "shred",
+      "truncate",
+      "shutdown",
+      "reboot",
+      "halt",
+      "poweroff",
+      "kill",
+      "killall",
+      "pkill",
+      "chmod",
+      "chown",
+      "mv",
+      "sudo",
+      "doas"
+    ]);
+    DESTRUCTIVE_GIT_SUBCOMMANDS = /* @__PURE__ */ new Set([
+      "reset",
+      "clean",
+      "push",
+      "rebase",
+      "filter-branch"
+    ]);
+    ALWAYS_ALLOW = /* @__PURE__ */ new Set(["read_file", "grep", "glob"]);
+  }
+});
+
+// src/tools/paths.ts
+import { resolve, relative as relative2, isAbsolute, sep, join as join5 } from "path";
+import { homedir as homedir4 } from "os";
 function isUnder(parent, child) {
   const rel = relative2(parent, child);
   return rel === "" || !rel.startsWith(".." + sep) && rel !== ".." && !isAbsolute(rel);
@@ -557,7 +691,7 @@ var SPILL_DIR;
 var init_paths = __esm({
   "src/tools/paths.ts"() {
     "use strict";
-    SPILL_DIR = resolve(join4(homedir3(), ".miii", "output"));
+    SPILL_DIR = resolve(join5(homedir4(), ".miii", "output"));
   }
 });
 
@@ -595,7 +729,7 @@ var init_verifyHint = __esm({
 });
 
 // src/tools/edit_file.ts
-import { readFileSync as readFileSync3, writeFileSync as writeFileSync3 } from "fs";
+import { readFileSync as readFileSync4, writeFileSync as writeFileSync4 } from "fs";
 function similarity(a, b) {
   const x = a.trim();
   const y = b.trim();
@@ -684,7 +818,7 @@ var init_edit_file = __esm({
             };
           }
           const abs = confinePath(path);
-          const src = readFileSync3(abs, "utf-8");
+          const src = readFileSync4(abs, "utf-8");
           const first = src.indexOf(old_str);
           if (first === -1) {
             if (replace_all !== true) {
@@ -692,7 +826,7 @@ var init_edit_file = __esm({
               if (fuzzy) {
                 const [s, e] = fuzzy;
                 const out2 = src.slice(0, s) + new_str + src.slice(e);
-                writeFileSync3(abs, out2, "utf-8");
+                writeFileSync4(abs, out2, "utf-8");
                 return { content: `Edited ${path} (whitespace-tolerant match).${verifyHint(path)}` };
               }
             }
@@ -707,7 +841,7 @@ var init_edit_file = __esm({
           }
           const out = all ? src.split(old_str).join(new_str) : src.slice(0, first) + new_str + src.slice(first + old_str.length);
           const n = all ? src.split(old_str).length - 1 : 1;
-          writeFileSync3(abs, out, "utf-8");
+          writeFileSync4(abs, out, "utf-8");
           return { content: `Edited ${path}${all ? ` (${n} occurrences)` : ""}.${verifyHint(path)}` };
         } catch (err) {
           return { content: err instanceof Error ? err.message : String(err), is_error: true };
@@ -718,7 +852,7 @@ var init_edit_file = __esm({
 });
 
 // src/tools/read_file.ts
-import { readFileSync as readFileSync4 } from "fs";
+import { readFileSync as readFileSync5 } from "fs";
 function numbered(lines, start) {
   const width = String(start + lines.length - 1).length;
   return lines.map((l, i) => `${String(start + i).padStart(width, " ")}	${l}`).join("\n");
@@ -743,7 +877,7 @@ var init_read_file = __esm({
       handler: ({ path, offset, limit }) => {
         try {
           const MAX_CHARS = 2e5;
-          const buf = readFileSync4(confinePath(path));
+          const buf = readFileSync5(confinePath(path));
           if (buf.subarray(0, 8e3).includes(0)) {
             return { content: `${path} looks binary (${buf.length} bytes); not reading as text.`, is_error: true };
           }
@@ -774,7 +908,7 @@ var init_read_file = __esm({
 });
 
 // src/tools/write_file.ts
-import { writeFileSync as writeFileSync4, mkdirSync as mkdirSync3 } from "fs";
+import { writeFileSync as writeFileSync5, mkdirSync as mkdirSync4 } from "fs";
 import { dirname } from "path";
 var write_file;
 var init_write_file = __esm({
@@ -796,8 +930,8 @@ var init_write_file = __esm({
       handler: ({ path, content }) => {
         try {
           const abs = confinePath(path);
-          mkdirSync3(dirname(abs), { recursive: true });
-          writeFileSync4(abs, content, "utf-8");
+          mkdirSync4(dirname(abs), { recursive: true });
+          writeFileSync5(abs, content, "utf-8");
           return { content: `Wrote ${path} (${content.length} bytes).${verifyHint(path)}` };
         } catch (err) {
           return { content: err instanceof Error ? err.message : String(err), is_error: true };
@@ -808,21 +942,21 @@ var init_write_file = __esm({
 });
 
 // src/tools/spill.ts
-import { writeFileSync as writeFileSync5, mkdirSync as mkdirSync4, rmSync as rmSync2, readdirSync as readdirSync3, statSync } from "fs";
-import { join as join5 } from "path";
-import { homedir as homedir4 } from "os";
+import { writeFileSync as writeFileSync6, mkdirSync as mkdirSync5, rmSync as rmSync2, readdirSync as readdirSync3, statSync } from "fs";
+import { join as join6 } from "path";
+import { homedir as homedir5 } from "os";
 import { randomBytes } from "crypto";
 function ensureDir() {
-  mkdirSync4(OUTPUT_DIR, { recursive: true });
+  mkdirSync5(OUTPUT_DIR, { recursive: true });
   return OUTPUT_DIR;
 }
 function spillIfLarge(full, label = "output", budget = INLINE_BUDGET) {
   if (full.length <= budget) return full;
   const id = randomBytes(6).toString("hex");
-  const file = join5(ensureDir(), `${id}.txt`);
+  const file = join6(ensureDir(), `${id}.txt`);
   let path = file;
   try {
-    writeFileSync5(file, full, "utf-8");
+    writeFileSync6(file, full, "utf-8");
   } catch {
     path = "";
   }
@@ -837,7 +971,7 @@ function cleanupSpill(maxAgeMs = 24 * 60 * 60 * 1e3) {
   try {
     const now = Date.now();
     for (const name of readdirSync3(OUTPUT_DIR)) {
-      const f = join5(OUTPUT_DIR, name);
+      const f = join6(OUTPUT_DIR, name);
       try {
         if (now - statSync(f).mtimeMs > maxAgeMs) rmSync2(f, { force: true });
       } catch {
@@ -850,7 +984,7 @@ var OUTPUT_DIR, INLINE_BUDGET, HEAD_FRACTION;
 var init_spill = __esm({
   "src/tools/spill.ts"() {
     "use strict";
-    OUTPUT_DIR = join5(homedir4(), ".miii", "output");
+    OUTPUT_DIR = join6(homedir5(), ".miii", "output");
     INLINE_BUDGET = 1e4;
     HEAD_FRACTION = 0.3;
   }
@@ -870,7 +1004,7 @@ var init_run_bash = __esm({
         type: "object",
         properties: {
           command: { type: "string", description: "Shell command to run" },
-          timeout_ms: { type: "number", description: "Timeout in ms (default 30000)" }
+          timeout_ms: { type: "number", description: "Timeout in ms (default 120000). Raise it for long builds/test suites." }
         },
         required: ["command"]
       },
@@ -880,7 +1014,7 @@ var init_run_bash = __esm({
           const shell = isWin ? "cmd" : "bash";
           const shellArgs = isWin ? ["/c", command] : ["-c", command];
           const { stdout, stderr, exitCode } = await execa(shell, shellArgs, {
-            timeout: timeout_ms ?? 3e4,
+            timeout: timeout_ms ?? 12e4,
             reject: false,
             all: false
           });
@@ -1187,14 +1321,14 @@ var init_validate = __esm({
 });
 
 // src/prompt/context.ts
-import { existsSync as existsSync3, readFileSync as readFileSync5, statSync as statSync3 } from "fs";
-import { dirname as dirname2, join as join6 } from "path";
+import { existsSync as existsSync4, readFileSync as readFileSync6, statSync as statSync3 } from "fs";
+import { dirname as dirname2, join as join7 } from "path";
 function findContextFile(cwd) {
   let dir = cwd;
   for (; ; ) {
-    const candidate = join6(dir, CONTEXT_FILENAME);
-    if (existsSync3(candidate)) return candidate;
-    if (existsSync3(join6(dir, ".git"))) return null;
+    const candidate = join7(dir, CONTEXT_FILENAME);
+    if (existsSync4(candidate)) return candidate;
+    if (existsSync4(join7(dir, ".git"))) return null;
     const parent = dirname2(dir);
     if (parent === dir) return null;
     dir = parent;
@@ -1205,7 +1339,7 @@ function loadProjectContext(cwd) {
   if (!source) return EMPTY;
   try {
     if (statSync3(source).size === 0) return { ...EMPTY, source };
-    const raw = readFileSync5(source, "utf8");
+    const raw = readFileSync6(source, "utf8");
     if (Buffer.byteLength(raw, "utf8") > MAX_CONTEXT_BYTES) {
       const clipped = Buffer.from(raw, "utf8").subarray(0, MAX_CONTEXT_BYTES).toString("utf8");
       return { content: clipped, source, truncated: true };
@@ -1311,6 +1445,7 @@ Ask in a numbered list. One round of questions per turn. Then wait.
 - After a tool result, move directly to the next tool call or the final answer. Do not restate what the previous tool did.
 - Every tool call MUST carry a complete, valid arguments object: all required fields present, correct types, valid JSON. Never emit a call with empty, partial, or placeholder arguments.
 - WRONG (leaks as text, nothing runs): writing \`call:some_tool{"foo":"bar"}\` or a fenced JSON block in your reply. RIGHT: emit it as a native function call with a full arguments object.
+- Batch independent tool calls in a SINGLE turn \u2014 parallel, not serial. If two reads, greps, or searches do not depend on each other's output, emit them together. Only serialize when a later call needs an earlier result.
 
 # Tools
 You have access to the following tools. Call them via the function-calling interface.
@@ -1320,14 +1455,33 @@ ${toolLines}
 - When you need to act on the filesystem or run a command, emit a tool call.
 - After each tool result, decide: more tool calls, or a final plain-text answer.
 - Stop emitting tool calls when GOAL is met. Reply with a concise plain-text final message confirming CRITERION is satisfied.
+- After the work is done, always close by asking the user what they want to do next \u2014 a brief, specific prompt (offer the most likely follow-ups when obvious). One line, no filler.
 
 # Rules
 - Always read a file before updating it. Never edit, overwrite, or create-over a file you have not read first this turn.
 - Prefer editing existing files over creating new ones.
+- To change an existing file, use edit_file with a small, targeted old_str/new_str diff \u2014 never rewrite the whole file with write_file. Reserve write_file for brand-new files or small ones. A full-file write_file on a large file risks getting cut off at the output token limit mid-write; a targeted edit_file stays small and avoids that.
+- When a new file's content is large, create it with write_file for the first portion, then append the rest with successive edit_file calls. Keep every write small.
 - For edit_file, make old_str unique by including surrounding context, or set replace_all to change every occurrence.
 - Never invent file paths. Read, glob, or grep before editing.
 - No empty filler or robotic boilerplate. A brief, genuine warm touch (see Tone and voice) is welcome; hollow pleasantries and reflexive apologies are not.
 
+# Scope discipline
+- Do ONLY what the user asked. No unrequested refactors, renames, reformatting, or "while I'm here" edits.
+- If you spot an unrelated issue worth fixing, mention it in your final message \u2014 do not fix it unprompted.
+- Touch the fewest files needed. A one-line request gets a one-line change, not a redesign.
+
+# Secrets and safety
+- Never print, log, or echo secrets, API keys, tokens, passwords, or \`.env\` values. Redact them if you must reference one.
+- Never write credentials into source, commits, or output. If a secret is needed, read it from the environment or config.
+- Do not exfiltrate file contents to external services without the user asking.
+
+# Git and commits
+- Do NOT commit, push, or create branches/PRs unless the user explicitly asks.
+- When asked to commit: never commit on the main branch \u2014 branch first. Stage only files relevant to the change; never blanket \`git add -A\` without checking what it sweeps in.
+- Write a concise commit message stating what changed and why. Do not add credentials or generated noise.
+- Never force-push, rebase shared history, or run destructive git commands without explicit confirmation.
+
 # Context discipline
 - read_file returns line numbers and accepts offset/limit. For large files, grep or glob to the relevant region first, then read only that range with offset/limit. Do not read a whole large file when you need a few functions \u2014 it wastes the context window.
 - Reference code by the line numbers read_file returns.
@@ -1349,94 +1503,6 @@ var init_system = __esm({
   }
 });
 
-// src/permissions/policy.ts
-import { readFileSync as readFileSync6, writeFileSync as writeFileSync6, mkdirSync as mkdirSync5, existsSync as existsSync4, renameSync } from "fs";
-import { join as join7 } from "path";
-import { homedir as homedir5 } from "os";
-function loadRules() {
-  if (!existsSync4(RULES_PATH)) return [];
-  try {
-    const data = JSON.parse(readFileSync6(RULES_PATH, "utf-8"));
-    return Array.isArray(data.rules) ? data.rules : [];
-  } catch {
-    return [];
-  }
-}
-function saveRules(rules) {
-  mkdirSync5(RULES_DIR, { recursive: true });
-  const tmp = RULES_PATH + ".tmp";
-  writeFileSync6(tmp, JSON.stringify({ rules }, null, 2), "utf-8");
-  renameSync(tmp, RULES_PATH);
-}
-function addRule(tool, pattern) {
-  const rules = loadRules();
-  if (rules.some((r) => r.tool === tool && r.pattern === pattern)) return;
-  rules.push({ tool, pattern });
-  saveRules(rules);
-}
-function subjectFor(toolName, input) {
-  const obj = input ?? {};
-  if (toolName === "run_bash") return typeof obj.command === "string" ? obj.command : "";
-  if (typeof obj.path === "string") return obj.path;
-  return "";
-}
-function generalizeCommand(command) {
-  const tokens = command.trim().split(/\s+/);
-  if (tokens.length === 0 || tokens[0] === "") return command;
-  const prog = tokens[0];
-  const prefixLen = WRAPPER_PROGRAMS.has(prog) && tokens.length > 1 ? 2 : 1;
-  const prefix = tokens.slice(0, prefixLen).join(" ");
-  return `${prefix} *`;
-}
-function patternToPersist(toolName, subject) {
-  return toolName === "run_bash" ? generalizeCommand(subject) : subject;
-}
-function globToRegExp(glob2) {
-  const escaped = glob2.replace(/[.+^${}()|[\]\\]/g, "\\$&");
-  const pattern = escaped.replace(/\*/g, ".*").replace(/\?/g, ".");
-  return new RegExp(`^${pattern}$`);
-}
-function matches(rule, toolName, subject) {
-  if (rule.tool !== toolName) return false;
-  try {
-    return globToRegExp(rule.pattern).test(subject);
-  } catch {
-    return false;
-  }
-}
-async function check(toolName, input, ctx) {
-  if (ALWAYS_ALLOW.has(toolName)) return "allow";
-  const subject = subjectFor(toolName, input);
-  const rules = loadRules();
-  if (rules.some((r) => matches(r, toolName, subject))) return "allow";
-  const answer = await ctx.ask(toolName, input);
-  if (answer === "no") return "deny";
-  if (answer === "always") addRule(toolName, patternToPersist(toolName, subject));
-  return "allow";
-}
-var RULES_DIR, RULES_PATH, WRAPPER_PROGRAMS, ALWAYS_ALLOW;
-var init_policy = __esm({
-  "src/permissions/policy.ts"() {
-    "use strict";
-    RULES_DIR = join7(homedir5(), ".miii");
-    RULES_PATH = join7(RULES_DIR, "permissions.json");
-    WRAPPER_PROGRAMS = /* @__PURE__ */ new Set([
-      "npm",
-      "npx",
-      "pnpm",
-      "yarn",
-      "brew",
-      "pip",
-      "pip3",
-      "cargo",
-      "docker",
-      "kubectl",
-      "go"
-    ]);
-    ALWAYS_ALLOW = /* @__PURE__ */ new Set(["read_file", "grep", "glob"]);
-  }
-});
-
 // src/agent/adapter.ts
 function mintToolUseId() {
   const rand = Math.random().toString(36).slice(2, 14);
@@ -1718,7 +1784,10 @@ async function* runAgent(opts) {
   const system = buildSystemPrompt(TOOLS, cwd, loadProjectContext(cwd));
   const ollamaTools = toOllamaTools(TOOLS);
   const toolNames = TOOLS.map((t) => t.name);
-  const effort = EFFORT_OPTIONS[loadConfig().effort ?? "medium"];
+  const cfg = loadConfig();
+  const effort = EFFORT_OPTIONS[cfg.effort ?? "medium"];
+  const ctxCap = cfg.numCtxCap && cfg.numCtxCap > 0 ? cfg.numCtxCap : DEFAULT_NUM_CTX_CAP;
+  const cappedCtx = typeof num_ctx === "number" && num_ctx > 0 ? Math.min(num_ctx, ctxCap) : void 0;
   const history = [
     ...opts.history,
     {
@@ -1733,6 +1802,7 @@ async function* runAgent(opts) {
   let repeatCount = 0;
   let leakNudges = 0;
   const seenPaths = /* @__PURE__ */ new Set();
+  let endedCleanly = false;
   for (let turn = 0; turn < MAX_TURNS; turn++) {
     let text = "";
     let tool_calls;
@@ -1745,7 +1815,7 @@ async function* runAgent(opts) {
     const composedSignal = signal ? AbortSignal.any ? AbortSignal.any([signal, ac.signal]) : ac.signal : ac.signal;
     if (signal) signal.addEventListener("abort", () => ac.abort(), { once: true });
     try {
-      for await (const chunk of chat3(model, toOllamaMessages(history, system), ollamaTools, { signal: composedSignal, num_ctx, num_predict: effort.num_predict, temperature: effort.temperature })) {
+      for await (const chunk of chat3(model, toOllamaMessages(history, system), ollamaTools, { signal: composedSignal, num_ctx: cappedCtx, num_predict: effort.num_predict, temperature: effort.temperature })) {
         if (signal?.aborted) break;
         if (chunk.content) {
           text += chunk.content;
@@ -1801,6 +1871,23 @@ async function* runAgent(opts) {
     }
     const blocks = blocksFromOllama(text, tool_calls, toolNames);
     const tool_uses = blocks.filter((b) => b.type === "tool_use");
+    if (tool_uses.length > 0 && !truncated) {
+      const sig = JSON.stringify(
+        blocks.map(
+          (b) => b.type === "tool_use" ? { t: "u", n: b.name, i: b.input } : b.type === "text" ? { t: "t", x: b.text.trim() } : b
+        )
+      );
+      if (sig === lastAssistantSig) {
+        repeatCount++;
+        if (repeatCount >= 2) {
+          yield { type: "error", message: "Agent loop detected: assistant produced identical output 3 turns in a row" };
+          return history;
+        }
+      } else {
+        repeatCount = 0;
+        lastAssistantSig = sig;
+      }
+    }
     history.push({ role: "assistant", content: blocks });
     if (truncated && tool_uses.length > 0) {
       const results2 = tool_uses.map((use) => ({
@@ -1826,24 +1913,10 @@ async function* runAgent(opts) {
         yield { type: "turn-end", stop_reason: "tool_use" };
         continue;
       }
+      endedCleanly = true;
       yield { type: "turn-end", stop_reason: "end_turn" };
       break;
     }
-    const sig = JSON.stringify(
-      blocks.map(
-        (b) => b.type === "tool_use" ? { t: "u", n: b.name, i: b.input } : b.type === "text" ? { t: "t", x: b.text.trim() } : b
-      )
-    );
-    if (sig === lastAssistantSig) {
-      repeatCount++;
-      if (repeatCount >= 2) {
-        yield { type: "error", message: "Agent loop detected: assistant produced identical output 3 turns in a row" };
-        return history;
-      }
-    } else {
-      repeatCount = 0;
-      lastAssistantSig = sig;
-    }
     for (const u of tool_uses) yield { type: "tool-use", block: u };
     const results = [];
     for (const use of tool_uses) {
@@ -1930,6 +2003,12 @@ async function* runAgent(opts) {
     history.push({ role: "user", content: results });
     yield { type: "turn-end", stop_reason: "tool_use" };
   }
+  if (!endedCleanly) {
+    yield {
+      type: "error",
+      message: `Stopped after ${MAX_TURNS} tool-use turns \u2014 the task may be incomplete. Send another message to continue where it left off.`
+    };
+  }
   yield { type: "done", prompt_tokens: promptTokens, eval_tokens: evalTokens };
   return history;
 }
@@ -2239,7 +2318,7 @@ var InputBar = memo(function InputBar2({ input, caret, disabled, processingLabel
   const [frame, setFrame] = useState(0);
   useEffect(() => {
     if (!disabled) return;
-    const t = setInterval(() => setFrame((f) => (f + 1) % SPIN.length), 150);
+    const t = setInterval(() => setFrame((f) => (f + 1) % SPIN.length), 200);
     return () => clearInterval(t);
   }, [disabled]);
   return /* @__PURE__ */ jsx2(
@@ -2277,7 +2356,7 @@ var InputBar = memo(function InputBar2({ input, caret, disabled, processingLabel
 // src/ui/ModelsView.tsx
 import { Box as Box3, Text as Text3 } from "ink";
 import { jsx as jsx3, jsxs as jsxs3 } from "react/jsx-runtime";
-function ModelsView({ models, cursor, model, host, provider, effort, query, requireSelection }) {
+function ModelsView({ models, cursor, model, host, provider, providerType, effort, query, requireSelection }) {
   return /* @__PURE__ */ jsxs3(Box3, { flexDirection: "column", marginLeft: 2, children: [
     /* @__PURE__ */ jsxs3(Box3, { flexDirection: "column", marginBottom: 1, children: [
       /* @__PURE__ */ jsxs3(Text3, { wrap: "truncate", children: [
@@ -2296,7 +2375,10 @@ function ModelsView({ models, cursor, model, host, provider, effort, query, requ
       ] })
     ] }),
     /* @__PURE__ */ jsx3(Text3, { dimColor: true, children: "select model" }),
-    /* @__PURE__ */ jsx3(Box3, { marginTop: 1, flexDirection: "column", borderStyle: "round", borderColor: "gray", paddingX: 1, children: models.length === 0 ? /* @__PURE__ */ jsx3(Text3, { dimColor: true, children: query ? `no models match "${query}"` : provider === "lmstudio" ? "no models. load a model in LM Studio and start the server." : "no models found." }) : models.map((m, i) => {
+    /* @__PURE__ */ jsx3(Box3, { marginTop: 1, flexDirection: "column", borderStyle: "round", borderColor: "gray", paddingX: 1, children: models.length === 0 ? query ? /* @__PURE__ */ jsx3(Text3, { dimColor: true, children: `no models match "${query}"` }) : provider === "lmstudio" ? /* @__PURE__ */ jsx3(Text3, { dimColor: true, children: "no models. load a model in LM Studio and start the server." }) : providerType === "ollama" ? /* @__PURE__ */ jsxs3(Box3, { flexDirection: "column", children: [
+      /* @__PURE__ */ jsx3(Text3, { dimColor: true, children: "no models installed. pull one, then relaunch:" }),
+      /* @__PURE__ */ jsx3(Text3, { color: "cyan", children: "  ollama pull qwen2.5-coder:14b" })
+    ] }) : /* @__PURE__ */ jsx3(Text3, { dimColor: true, children: `no models found at ${host}. make sure the server is running with a model loaded.` }) : models.map((m, i) => {
       const sel = i === cursor;
       return /* @__PURE__ */ jsxs3(Text3, { wrap: "truncate", color: sel ? "blue" : void 0, dimColor: !sel, children: [
         sel ? "\u276F " : "  ",
@@ -3285,14 +3367,26 @@ function clipTail(rendered, max) {
   if (lines.length <= max) return { text: rendered, clipped: 0 };
   return { text: lines.slice(-max).join("\n"), clipped: lines.length - max };
 }
+var ANSI_RE = /\x1b\[[0-9;]*m/g;
+function stripAnsi(s) {
+  return s.replace(ANSI_RE, "");
+}
+function visualHeight(text, width) {
+  const w = Math.max(1, width);
+  let rows = 0;
+  for (const line of text.split("\n")) {
+    rows += Math.max(1, Math.ceil(stripAnsi(line).length / w));
+  }
+  return rows;
+}
 function clipTailVisual(content, maxRows, width) {
   const w = Math.max(1, width);
   const lines = content.split("\n");
-  const visualRows = (line) => Math.max(1, Math.ceil(line.length / w));
+  const visualRows2 = (line) => Math.max(1, Math.ceil(stripAnsi(line).length / w));
   let rows = 0;
   let start = lines.length;
   for (let i = lines.length - 1; i >= 0; i--) {
-    const h = visualRows(lines[i]);
+    const h = visualRows2(lines[i]);
     if (rows + h > maxRows && start < lines.length) break;
     rows += h;
     start = i;
@@ -3305,9 +3399,20 @@ function liveFrameRows() {
   return Math.max(6, rows - 8);
 }
 var COLLAPSED_LINES = 3;
+function visualRows(text, width, cap) {
+  const w = Math.max(1, width);
+  let rows = 0;
+  const lines = text.split("\n");
+  for (const line of lines) {
+    rows += Math.max(1, Math.ceil(line.length / w));
+    if (rows >= cap) return cap;
+  }
+  return rows;
+}
 function estimateToolRows(use, result) {
   const input = use.input ?? {};
   const noErr = !result?.is_error;
+  const w = contentWidth();
   if (use.name === "write_file" && noErr) {
     const total = countLines(String(input.content ?? ""));
     const shown = Math.min(total, COLLAPSED_LINES);
@@ -3323,10 +3428,10 @@ function estimateToolRows(use, result) {
     const lines = (result.content ?? "").split("\n");
     const multi = (use.name === "run_bash" || use.name === "grep" || use.name === "glob" || result.is_error) && lines.length > 1;
     if (multi) {
-      const shown = Math.min(lines.length, COLLAPSED_LINES);
-      rows += 1 + shown + (lines.length > shown ? 1 : 0);
+      const shownLines = lines.slice(0, COLLAPSED_LINES).join("\n");
+      rows += 1 + visualRows(shownLines, w, COLLAPSED_LINES * 4) + (lines.length > COLLAPSED_LINES ? 1 : 0);
     } else {
-      rows += 1;
+      rows += visualRows(lines[0] ?? "", w, 4);
     }
   }
   return rows;
@@ -3338,6 +3443,7 @@ function contentWidth() {
 // src/ui/ThinkingBlock.tsx
 import { jsx as jsx8, jsxs as jsxs8 } from "react/jsx-runtime";
 var FRAMES = ["\u280B", "\u2819", "\u2839", "\u2838", "\u283C", "\u2834", "\u2826", "\u2827", "\u2807", "\u280F"];
+var CHALK = "#c9c7c0";
 var globalThinkingVisible = false;
 var listeners = /* @__PURE__ */ new Set();
 function toggleThinkingVisible() {
@@ -3359,16 +3465,17 @@ function ThinkingBlock({ content }) {
   const [frame, setFrame] = useState2(0);
   const visible = useThinkingVisible();
   useEffect2(() => {
-    const t = setInterval(() => setFrame((f) => (f + 1) % FRAMES.length), 80);
+    const t = setInterval(() => setFrame((f) => (f + 1) % FRAMES.length), 100);
     return () => clearInterval(t);
   }, []);
+  const label = "thinking";
   return /* @__PURE__ */ jsxs8(Box8, { flexDirection: "column", marginLeft: 2, marginBottom: 1, children: [
     /* @__PURE__ */ jsxs8(Box8, { children: [
-      /* @__PURE__ */ jsxs8(Text8, { color: "blue", children: [
+      /* @__PURE__ */ jsxs8(Text8, { color: CHALK, children: [
         FRAMES[frame],
         " "
       ] }),
-      /* @__PURE__ */ jsx8(Text8, { dimColor: true, italic: true, children: "thinking\u2026" }),
+      /* @__PURE__ */ jsx8(Text8, { color: CHALK, italic: true, children: label }),
       /* @__PURE__ */ jsxs8(Text8, { dimColor: true, children: [
         " \xB7 ctrl+t to ",
         visible ? "hide" : "show",
@@ -3678,6 +3785,7 @@ var AssistantMessage = memo2(function AssistantMessage2({ msg }) {
 
 // src/ui/PermissionPrompt.tsx
 import { Box as Box11, Text as Text11 } from "ink";
+init_policy();
 import { jsx as jsx11, jsxs as jsxs11 } from "react/jsx-runtime";
 function summarizeInput(input) {
   if (!input || typeof input !== "object") return "";
@@ -3699,9 +3807,10 @@ function summarizeInput(input) {
 }
 function PermissionPrompt({ req, cursor }) {
   const label = TOOL_LABEL[req.toolName] ?? req.toolName;
+  const rule = patternToPersist(req.toolName, subjectFor(req.toolName, req.input));
   const options = [
     { label: "Yes", key: "yes" },
-    { label: "Yes, don't ask again for this", key: "always" },
+    { label: rule ? `Yes, don't ask again for ${rule}` : "Yes, don't ask again for this", key: "always" },
     { label: "No", key: "no" }
   ];
   const summary = summarizeInput(req.input);
@@ -3735,7 +3844,8 @@ function ChatView({
   permissionCursor = 0,
   activeToolUses,
   activeToolResults,
-  header
+  header,
+  logEpoch = 0
 }) {
   const empty = messages.length === 0 && !streaming && !thinking && !pendingPermission && !error;
   const log = [];
@@ -3750,8 +3860,12 @@ function ChatView({
   let streamNode = null;
   let streamRows = 0;
   if (streaming && streamingContent) {
-    const { text, clipped } = clipTail(renderMarkdownStreaming(streamingContent), liveBudget);
-    streamRows = text.split("\n").length + (clipped > 0 ? 1 : 0);
+    const raw = clipTail(streamingContent, liveBudget);
+    const width = contentWidth();
+    const rendered = clipTailVisual(renderMarkdownStreaming(raw.text), liveBudget, width);
+    const text = rendered.text;
+    const clipped = raw.clipped + rendered.clipped;
+    streamRows = visualHeight(text, width) + (clipped > 0 ? 1 : 0);
     streamNode = /* @__PURE__ */ jsxs12(Box12, { flexDirection: "column", marginBottom: 1, children: [
       clipped > 0 && /* @__PURE__ */ jsx12(Text12, { dimColor: true, children: `\u2191 ${clipped} more line${clipped === 1 ? "" : "s"} above \u2014 streaming\u2026` }),
       /* @__PURE__ */ jsxs12(Box12, { flexDirection: "row", children: [
@@ -3781,7 +3895,7 @@ function ChatView({
     ] });
   }
   return /* @__PURE__ */ jsxs12(Fragment2, { children: [
-    /* @__PURE__ */ jsx12(Static, { items: log, children: (item) => item.key === "header" ? /* @__PURE__ */ jsx12(Box12, { children: item.node }, item.key) : /* @__PURE__ */ jsx12(Box12, { marginLeft: 1, children: item.node }, item.key) }),
+    /* @__PURE__ */ jsx12(Static, { items: log, children: (item) => item.key === "header" ? /* @__PURE__ */ jsx12(Box12, { children: item.node }, item.key) : /* @__PURE__ */ jsx12(Box12, { marginLeft: 1, children: item.node }, item.key) }, logEpoch),
     /* @__PURE__ */ jsxs12(Box12, { flexDirection: "column", marginLeft: 1, marginBottom: 1, children: [
       empty && /* @__PURE__ */ jsxs12(Box12, { flexDirection: "column", marginBottom: 1, children: [
         /* @__PURE__ */ jsx12(Text12, { dimColor: true, children: EMPTY_STATE_TITLE }),
@@ -3949,8 +4063,8 @@ function useAgentRunner(model, activeCtx) {
           case "turn-end": {
             flushStream(true);
             flushThink(true);
-            setStreaming(false);
             if (ev.stop_reason === "tool_use") {
+              setStreaming(false);
               flushTurn(null);
               setThinking(true);
               thinkingAcc = "";
@@ -4222,6 +4336,7 @@ function useKeyboard(opts) {
     sessions,
     setSessions,
     setNotice,
+    setLogEpoch,
     switchProvider
   } = opts;
   const {
@@ -4255,6 +4370,7 @@ function useKeyboard(opts) {
     setError(null);
     setNotice(null);
     clearPasteStore();
+    setLogEpoch((n) => n + 1);
   }
   const effort = cfg.effort ?? "medium";
   useInput((char, key) => {
@@ -4713,6 +4829,7 @@ function App() {
   sessionIdRef.current = sessionId;
   const [sessions, setSessions] = useState5([]);
   const [notice, setNotice] = useState5(null);
+  const [logEpoch, setLogEpoch] = useState5(0);
   const [input, setInput] = useState5("");
   const [caret, setCaret] = useState5(0);
   const [paletteCursor, setPaletteCursor] = useState5(0);
@@ -4833,6 +4950,7 @@ function App() {
     sessions,
     setSessions,
     setNotice,
+    setLogEpoch,
     switchProvider
   });
   const effort = cfg.effort ?? "medium";
@@ -4844,7 +4962,7 @@ function App() {
     return Math.round(used / activeCtx * 100);
   })();
   return /* @__PURE__ */ jsxs13(Box13, { flexDirection: "column", paddingX: 1, children: [
-    state !== "ready" && /* @__PURE__ */ jsx13(WelcomeBlock, { model: cfg.model, activeCtx, effort, cwd, error: agent.error, updateAvailable, updateStatus }),
+    state !== "ready" && state !== "sessions" && state !== "models" && /* @__PURE__ */ jsx13(WelcomeBlock, { model: cfg.model, activeCtx, effort, cwd, error: agent.error, updateAvailable, updateStatus }),
     state === "loading" && !agent.error && /* @__PURE__ */ jsx13(Box13, { marginLeft: 2, marginBottom: 1, children: /* @__PURE__ */ jsx13(Text13, { dimColor: true, children: `connecting to ${provName}\u2026` }) }),
     agent.error && state !== "ready" && /* @__PURE__ */ jsx13(
       ChatView,
@@ -4856,7 +4974,7 @@ function App() {
         error: agent.error
       }
     ),
-    (state === "select-model" || state === "models") && /* @__PURE__ */ jsx13(
+    state === "select-model" && /* @__PURE__ */ jsx13(
       ModelsView,
       {
         models: filteredModels,
@@ -4864,9 +4982,10 @@ function App() {
         model: cfg.model,
         host: provEntry.baseUrl,
         provider: provName,
+        providerType: provEntry.type,
         effort,
         query: pickerQuery,
-        requireSelection: state === "select-model"
+        requireSelection: true
       }
     ),
     state === "providers" && /* @__PURE__ */ jsx13(
@@ -4878,8 +4997,7 @@ function App() {
         query: pickerQuery
       }
     ),
-    state === "sessions" && /* @__PURE__ */ jsx13(SessionsView, { sessions, cursor }),
-    state === "ready" && /* @__PURE__ */ jsxs13(Fragment3, { children: [
+    (state === "ready" || state === "sessions" || state === "models") && /* @__PURE__ */ jsxs13(Fragment3, { children: [
       notice && /* @__PURE__ */ jsx13(Box13, { marginLeft: 2, marginBottom: 1, children: /* @__PURE__ */ jsx13(Text13, { color: "green", children: `\u2713 ${notice}` }) }),
       /* @__PURE__ */ jsx13(
         ChatView,
@@ -4894,18 +5012,33 @@ function App() {
           permissionCursor: agent.permissionCursor,
           activeToolUses: agent.activeToolUses,
           activeToolResults: agent.activeToolResults,
-          header: /* @__PURE__ */ jsx13(WelcomeBlock, { model: cfg.model, activeCtx, effort, cwd })
+          header: /* @__PURE__ */ jsx13(WelcomeBlock, { model: cfg.model, activeCtx, effort, cwd }),
+          logEpoch
         }
       ),
-      input.startsWith("/") && /* @__PURE__ */ jsx13(CommandPalette, { filter: input, cursor: paletteCursor }),
-      contextWarning !== null && /* @__PURE__ */ jsx13(Box13, { marginLeft: 2, marginBottom: 1, children: /* @__PURE__ */ jsx13(Text13, { color: "yellow", children: `\u26A0 context ${contextWarning}% full \u2014 run /clear and start fresh` }) }),
-      !input.startsWith("/") && (() => {
+      state === "ready" && input.startsWith("/") && /* @__PURE__ */ jsx13(CommandPalette, { filter: input, cursor: paletteCursor }),
+      state === "ready" && contextWarning !== null && /* @__PURE__ */ jsx13(Box13, { marginLeft: 2, marginBottom: 1, children: /* @__PURE__ */ jsx13(Text13, { color: "yellow", children: `\u26A0 context ${contextWarning}% full \u2014 run /clear and start fresh` }) }),
+      state === "ready" && !input.startsWith("/") && (() => {
         const m = parseMention(input);
         if (!m) return null;
         return /* @__PURE__ */ jsx13(FilePicker, { matches: searchFiles(process.cwd(), m.query), cursor: filePickerCursor });
       })(),
-      /* @__PURE__ */ jsx13(InputBar, { input, caret, disabled: agent.busy, processingLabel: agent.processingLabel }),
-      !agent.busy && /* @__PURE__ */ jsx13(Box13, { marginLeft: 2, marginBottom: 1, children: /* @__PURE__ */ jsx13(Text13, { dimColor: true, children: providerDown ? "provider unavailable \u2014 /provider to switch \xB7 /models to pick a model" : "type / to see commands" }) }),
+      state === "ready" && /* @__PURE__ */ jsx13(InputBar, { input, caret, disabled: agent.busy, processingLabel: agent.processingLabel }),
+      state === "sessions" && /* @__PURE__ */ jsx13(SessionsView, { sessions, cursor }),
+      state === "models" && /* @__PURE__ */ jsx13(
+        ModelsView,
+        {
+          models: filteredModels,
+          cursor,
+          model: cfg.model,
+          host: provEntry.baseUrl,
+          provider: provName,
+          providerType: provEntry.type,
+          effort,
+          query: pickerQuery
+        }
+      ),
+      state === "ready" && !agent.busy && /* @__PURE__ */ jsx13(Box13, { marginLeft: 2, marginBottom: 1, children: /* @__PURE__ */ jsx13(Text13, { dimColor: true, children: providerDown ? "provider unavailable \u2014 /provider to switch \xB7 /models to pick a model" : "type / to see commands" }) }),
       updateAvailable && /* @__PURE__ */ jsx13(Box13, { marginLeft: 2, marginBottom: 1, children: /* @__PURE__ */ jsx13(Text13, { color: updateStatus === "failed" ? "red" : updateStatus === "installed" ? "green" : "yellow", children: updateBannerText(updateAvailable, updateStatus) }) })
     ] })
   ] });
@@ -4940,6 +5073,8 @@ if (cmd === "version" || cmd === "--version" || cmd === "-v") {
   const { runEval: runEval2 } = await Promise.resolve().then(() => (init_run(), run_exports));
   process.exit(await runEval2(rest));
 } else {
+  const cfgErr = configError();
+  if (cfgErr) console.error(cfgErr);
   process.on("exit", () => {
     if (process.stdout.isTTY) process.stdout.write("\x1B]2;\x07");
   });

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "miii-agent",
-  "version": "0.1.29",
-  "description": "Cursor / Claude Code, but local. An offline AI pair-programmer in your terminal, powered by Ollama. Private by default, free forever.",
+  "version": "0.1.31",
+  "description": "Local AI coding agent for your terminal — an open-source, offline alternative to Claude Code, Cursor, and GitHub Copilot, powered by Ollama. Private by default, free forever.",
   "type": "module",
   "bin": {
     "miii": "dist/cli.js"
@@ -52,7 +52,14 @@
     "pair-programming",
     "code-generation",
     "llama-cpp",
-    "lm-studio"
+    "lm-studio",
+    "copilot",
+    "copilot-alternative",
+    "cursor-alternative",
+    "claude-code",
+    "claude-code-alternative",
+    "developer-tools",
+    "qwen"
   ],
   "license": "MIT",
   "dependencies": {