nex-code 0.4.16 → 0.4.18

package/README.md

@@ -1,13 +1,16 @@
-```
- ██▄▄██   nex-code  v0.3.54
- █▀██▀█   devstral-2:123b  ·  /help
- ▀████▀
-```
+<h1 align="center">nex-code</h1>
+
+<p align="center">
+  <b>Run 400B+ open coding models on your codebase — without the hardware bill.</b><br>
+  Ollama Cloud first. OpenAI, Anthropic, and Gemini when you need them.
+</p>
+
+<p align="center">
+  <code>npx nex-code</code>
+</p>
 
 <p align="center">
-  <b>The open-source agentic coding CLI for Ollama Cloud.</b><br>
-  Use it in the terminal or install the built-in <b>VS Code extension</b> for a sidebar chat panel.<br>
-  Built and optimized for <a href="https://ollama.com">Ollama Cloud</a> — OpenAI, Anthropic, and Gemini are also supported.
+  <a href="https://github.com/hybridpicker/nex-code/stargazers">⭐ If this saves you time, a star helps others find it.</a>
 </p>
 
 <p align="center">
@@ -25,6 +28,14 @@
 
 ---
 
+```
+ ██▄▄██   nex-code  v0.3.54
+ █▀██▀█   devstral-2:123b  ·  /help
+ ▀████▀
+```
+
+---
+
 ## Demo
 
 https://github.com/user-attachments/assets/6a70525c-a1a4-4c6f-9784-176cb1becc15
@@ -65,6 +76,21 @@ npm update -g nex-code
 
 ---
 
+## vs. Claude Code & Gemini CLI
+
+| | nex-code | Claude Code | Gemini CLI |
+|---|---|---|---|
+| Free tier | ✅ Ollama Cloud flat-rate | ❌ subscription required | ⚠️ limited free quota |
+| Open models | ✅ devstral, Kimi K2, Qwen3 | ❌ Anthropic only | ❌ Google only |
+| Local Ollama | ✅ | ❌ | ❌ |
+| Multi-provider | ✅ swap with one env var | ❌ | ❌ |
+| VS Code sidebar | ✅ built-in, same install | ✅ | ❌ |
+| Startup time | ~100ms | ~2–4s | ~1–2s |
+| Runtime deps | 2 | heavy | heavy |
+| Infra tools | ✅ SSH, Docker, K8s built-in | ❌ | ❌ |
+
+---
+
 ## Why nex-code?
 
 **Ollama Cloud first.** nex-code is built and optimized for [Ollama Cloud](https://ollama.com) — the flat-rate platform that runs devstral, Kimi K2, Qwen3-Coder, and 47+ other open models. All behavioral tuning (loop detection, context compression, tool-call repair) is done against real Ollama Cloud sessions. Other providers (OpenAI, Anthropic, Gemini) work via the same interface but are not the primary target.
@@ -119,19 +145,18 @@ Rankings are based on nex-code's own `/benchmark` — 15 tool-calling tasks agai
 ### Flat-Rate / Pay-as-you-go
 
 <!-- nex-benchmark-start -->
-<!-- Updated: 2026-03-20 — run `/benchmark --discover` after new Ollama Cloud releases -->
+<!-- Updated: 2026-03-26 — run `/benchmark --discover` after new Ollama Cloud releases -->
 
-| Rank | Model                  | Score  | Avg Latency | Context | Best For                                              |
-| ---- | ---------------------- | ------ | ----------- | ------- | ----------------------------------------------------- |
-| 🥇   | `devstral-2:123b`      | **84** | 1.5s        | 131K    | Default — fastest + most reliable tool selection      |
-| 🥈   | `qwen3-coder:480b`     | 79     | 2.9s        | 131K    | Coding-heavy sessions, heavy sub-agents               |
-| 🥉   | `kimi-k2:1t`           | 79     | 2.7s        | 256K    | Large repos (>100K tokens)                            |
-| —    | `minimax-m2.7:cloud`   | 73     | 3.5s        | 200K    | Complex swarm / multi-agent sessions (Toolathon SOTA) |
-| —    | `devstral-small-2:24b` | 73     | 1.0s        | 131K    | Fast sub-agents, simple lookups                       |
+| Rank | Model | Score | Avg Latency | Context | Best For |
+|---|---|---|---|---|---|
+| 🥇 | `devstral-2:123b` | **82.5** | 1.7s | 131K | Default — fastest + most reliable tool selection |
+| 🥈 | `devstral-small-2:24b` | 75 | 3.1s | 131K | Fast sub-agents, simple lookups |
+| 🥉 | `qwen3-coder:480b` | 72.5 | 8.4s | 131K | Coding-heavy sessions, heavy sub-agents |
+| — | `kimi-k2:1t` | 67.5 | 6.6s | 256K | Large repos (>100K tokens) |
+| — | `minimax-m2.7:cloud` | 64.1 | 5.0s | 200K | Complex swarm / multi-agent sessions (Toolathon SOTA) |
 
 > Rankings are nex-code-specific: tool name accuracy, argument validity, schema compliance.
 > Toolathon (Minimax SOTA) measures different task types — run `/benchmark --discover` after model releases.
-
 <!-- nex-benchmark-end -->
 
 ### Recommended `.env` for Ollama Cloud (Flat-Rate)
@@ -344,7 +369,7 @@ nex-code --prompt-file /tmp/task.txt --yolo --json
 | `--max-turns <n>`          | Override the agentic loop iteration limit                                                                     |
 | `--model <spec>`           | Use a specific model (e.g. `anthropic:claude-sonnet-4-6`)                                                     |
 | `--debug`                  | Show internal diagnostic messages (compression, loop detection, guards)                                       |
-| `--auto-orchestrate`       | Automatically use the multi-agent orchestrator when ≥3 goals are detected (also: `NEX_AUTO_ORCHESTRATE=true`) |
+| `--no-auto-orchestrate`    | Disable auto-orchestration for multi-goal prompts (on by default; also: `NEX_AUTO_ORCHESTRATE=false`)         |
 | `--orchestrator-model <m>` | Model for decomposition/synthesis step (default: `kimi-k2.5`)                                                 |
 
 ---
@@ -690,10 +715,10 @@ deploy the latest build to prod
 The agent loop uses a bouncing-ball spinner (`● · · · ·` → `· ● · · ·` → …) during tool execution, then prints compact 1-line summaries:
 
 ```
-  ●     ▸ 3 tools: read_file, grep, edit_file
-  ✓ read_file src/app.js (45 lines)
-  ✓ grep TODO → 12 matches
-  ✗ edit_file src/x.js → old_text not found
+  ⏺ Read, Grep, Edit
+  ⎿  Read 45 lines from app.js
+  ⎿  12 matches "TODO"
+  ⎿  old_text not found
 ```
 
 After multi-step tasks, a résumé and context-aware follow-up suggestions are shown:
@@ -947,7 +972,7 @@ When the agent creates a task list, a **live animated display** replaces the sta
      ◻ Run tests
 ```
 
-- Bouncing-ball spinner (`●` ping-pong across 5 positions) with elapsed time display
+- Bouncing-ball spinner (⏺ ping-pong across 5 positions) with elapsed time display
 - Per-task status icons: `✔` done, `◼` in progress, `◻` pending, `✗` failed
 - Automatically pauses during text streaming and resumes during tool execution
 - Falls back to the static `/tasks` view when no live display is active
@@ -965,19 +990,19 @@ Spawn parallel sub-agents for independent tasks:
 
 For complex tasks with multiple independent goals (e.g. "fix all TypeScript errors in auth/, add tests for utils/, and update the README"), the orchestrator decomposes the prompt into parallel sub-tasks, runs dedicated sub-agents on each, and synthesizes the results.
 
-**Trigger:**
+**Auto-orchestration is on by default** for prompts with ≥3 goals.
 
 ```bash
-# One-off: pass flag on the CLI
-nex-code --auto-orchestrate --task "fix all type errors in src/, add JSDoc to utils/, update CHANGELOG"
+# Just use it — multi-goal prompts auto-decompose into parallel agents
+nex-code --task "fix all type errors in src/, add JSDoc to utils/, update CHANGELOG"
 
-# Headless with custom orchestrator model
-nex-code --auto-orchestrate --orchestrator-model kimi-k2.5 --task "..."
+# Custom orchestrator model
+nex-code --orchestrator-model kimi-k2.5 --task "..."
 
-# Always-on via env var
-NEX_AUTO_ORCHESTRATE=true nex-code
+# Disable auto-orchestration
+NEX_AUTO_ORCHESTRATE=false nex-code
 
-# Set minimum goal count before orchestrator activates (default: 3)
+# Lower the goal threshold (default: 3)
 NEX_ORCHESTRATE_THRESHOLD=2 nex-code
 ```
 
@@ -1013,7 +1038,7 @@ Suggested commit: fix: resolve auth type errors and add utility docs
 
 | Variable                    | Default | Description                                             |
 | --------------------------- | ------- | ------------------------------------------------------- |
-| `NEX_AUTO_ORCHESTRATE`      | `false` | Set to `true` to always use the orchestrator            |
+| `NEX_AUTO_ORCHESTRATE`      | `true`  | Set to `false` to disable auto-orchestration            |
 | `NEX_ORCHESTRATE_THRESHOLD` | `3`     | Minimum number of detected goals before auto-triggering |
 
 **Model roles in orchestration:**