0.5.40

Author: vava-nessa

changelog/v0.5.40.md

@@ -0,0 +1,4 @@
+# Changelog v0.5.40 - 2026-07-07
+
+### Fixed
+- Avoid sending `thinking: { type: 'disabled' }` to Cerebras, Mistral, Groq, and SambaNova during ping probes, preventing immediate 400 errors and redundant retry calls.

package.json

@@ -1,6 +1,6 @@
 {
   "name": "free-coding-models",
-  "version": "0.5.39",
+  "version": "0.5.40",
   "description": "Find the fastest coding LLM models in seconds — ping free models from multiple providers, pick the best one for OpenCode, Cursor, or any AI coding assistant.",
   "keywords": [
     "nvidia",

pi-extension/README.md

@@ -0,0 +1,94 @@
+# FCM-Pi — Pi Extension for free-coding-models
+
+A lightweight JavaScript extension for the **Pi coding agent (pi.dev)** that connects to **free-coding-models** (FCM). It runs round-trip pings and real-answer AI latency benchmarks on your configured free models, and automatically activates the most performant, usable model before your coding session starts.
+
+---
+
+## ✨ Features
+
+- **Auto-scan on session startup**: Runs a quick scan when Pi boots up, updating status dynamically in the terminal footer.
+- **Smart model selection**: Auto-configures Pi to use the highest-performing model based on a composite score:
+  - **60% SWE-bench score** (Coding capability)
+  - **20% Latency** (Network round-trip response speed)
+  - **10% TPS** (Throughput token speed)
+  - **10% Stability** (Success rates / uptime)
+- **Interactive selection (`/fcm`)**: Allows manually selecting from the top 10 best-performing models in real-time.
+- **Diagnostics (`/fcm-list` / `/fcm-status`)**: Displays ranked models in clean, structured terminal tables, and checks local daemon status.
+- **Smart Router Integration (`/fcm-router`)**: Routes requests through a local model-routing daemon with automatic failover fallback chains.
+
+---
+
+## 🚀 Quick Start
+
+### Prerequisites
+
+1. **free-coding-models** must be installed and configured:
+   ```bash
+   npm install -g free-coding-models
+   
+   # Run once to configure your API keys (Groq, Google AI Studio, Nvidia NIM, etc.)
+   free-coding-models
+   ```
+
+2. **Pi coding agent (pi.dev)** must be installed on your system.
+
+### Installation
+
+Install the extension directly inside the Pi shell or terminal:
+
+```bash
+pi install /Users/vava/Documents/GitHub/free-coding-models/pi-extension
+```
+
+*Or, if published to npm:*
+
+```bash
+pi install npm:fcm-pi
+```
+
+---
+
+## 📐 Architecture & Scan Strategies
+
+```
+               [Pi Session Starting]
+                         │
+                         ▼
+             Check: Is FCM Daemon active?
+             (http://localhost:19280)
+               /                  \
+             YES                  NO
+             /                      \
+    [Daemon Scan (~1s)]      [Direct Scan (~8-15s)]
+     Query cache & stats      Ping 30 candidate models
+     directly from daemon     Benchmark top 5 candidates
+             \                      /
+              \                    /
+             Filter: status === 'up'
+             Rank: Compute composite scores
+             Select: Write Pi configurations
+                         │
+                         ▼
+             [Pi Session Ready to Code!]
+```
+
+- **Daemon Mode**: Checks if the FCM background daemon is active. If so, fetches pre-cached latency and stats in under 1 second.
+- **Direct Mode (Fallback)**: Performs a parallel ping of the top 30 models from the FCM registry, then benchmarks the top 5 smartest models, completing in ~8-15 seconds.
+- **Config Persistence**: Modifies `~/.pi/agent/models.json` and `~/.pi/agent/settings.json` to store provider URLs and default model choices.
+
+---
+
+## ⌨️ Command Reference
+
+| Command | Action |
+|---------|--------|
+| `/fcm` | Re-runs diagnostic scan and presents interactive top 10 model list |
+| `/fcm-list` | Renders a styled table of the top 20 available models with SWE/Latency/TPS |
+| `/fcm-router` | Configures Pi to route completions through the FCM Smart Router daemon |
+| `/fcm-status` | Shows status diagnostics (active model, scan latency, daemon state) |
+
+---
+
+## 🧹 Telemetry & Security
+
+This extension does not collect or transmit API keys, source code, or personal data. It works entirely offline or connects only to your configured AI providers and local loopback address (`http://localhost:19280`).