free-coding-models 0.3.66 → 0.3.68

package/README.md

@@ -125,6 +125,94 @@ Use ⚡️ Command Palette! with **Ctrl+P**.
   <img src="https://img.shields.io/badge/USE_%E2%9A%A1%EF%B8%8F%20COMMAND%20PALETTE-CTRL%2BP-22c55e?style=for-the-badge" alt="Use ⚡️ Command Palette with Ctrl+P">
 </p>
 
+---
+
+## 🐳 Docker
+
+Run FCM without installing Node.js using the official Docker image:
+
+```bash
+# Quick start (daemon + web UI on port 19280)
+docker run -p 19280:19280 ghcr.io/vava-nessa/free-coding-models:latest
+
+# With an API key
+docker run -p 19280:19280 -e OPENROUTER_API_KEY=your_key ghcr.io/vava-nessa/free-coding-models:latest
+```
+
+Access the web dashboard at `http://localhost:19280/` and configure your coding tool to use `http://localhost:19280/v1` with model `fcm`.
+
+### Available Image Tags
+
+| Tag | Description |
+|-----|-------------|
+| `latest` | Most recent release |
+| `v{major}.{minor}.{patch}` | Specific version (e.g., `v0.3.70`) |
+| `v{major}.{minor}` | Minor version (e.g., `v0.3`) |
+| `v{major}` | Major version (e.g., `v0`) |
+
+### Environment Variables
+
+| Variable | Default | Description |
+|----------|---------|-------------|
+| `FCM_HOST` | `0.0.0.0` | Host to bind to (set `127.0.0.1` for localhost-only) |
+| `FCM_PORT` | `19280` | Port to listen on |
+| `FREE_CODING_MODELS_TELEMETRY` | `0` | Disable telemetry |
+
+Provider API keys (all optional):
+
+```bash
+docker run -p 19280:19280 \
+  -e NVIDIA_API_KEY=your_key \
+  -e GROQ_API_KEY=your_key \
+  -e OPENROUTER_API_KEY=your_key \
+  ghcr.io/vava-nessa/free-coding-models:latest
+```
+
+### Docker Compose
+
+Create a `docker-compose.yml`:
+
+```yaml
+version: '3.8'
+services:
+  fcm:
+    image: ghcr.io/vava-nessa/free-coding-models:latest
+    container_name: fcm
+    restart: unless-stopped
+    ports:
+      - "19280:19280"
+    environment:
+      FREE_CODING_MODELS_TELEMETRY: "0"
+      FCM_HOST: "0.0.0.0"
+      OPENROUTER_API_KEY: ${OPENROUTER_API_KEY:-}
+    volumes:
+      - fcm-data:/home/fcm
+volumes:
+  fcm-data:
+```
+
+Run with `docker-compose up -d`. API keys can be passed via a `.env` file or environment variables.
+
+### Troubleshooting
+
+**Container won't start:**
+- Check logs: `docker logs fcm`
+- Verify port 19280 is not in use: `docker ps | grep 19280`
+
+**Health check fails:**
+- Wait 30s for initial probe cycle
+- Verify API keys are valid: `docker exec fcm curl http://localhost:19280/health`
+
+**Cannot connect from host:**
+- Ensure `FCM_HOST=0.0.0.0` (default)
+- Check firewall allows localhost connections
+
+**Data persistence:**
+- Config is stored in Docker volume `fcm-data`
+- Recreate the volume with `docker-compose down -v` to reset
+
+---
+
 Need to fix contrast because your terminal theme is fighting the TUI? Press **`G`** at any time to cycle **Auto → Dark → Light**. The switch recolors the full interface live: table, Settings, Help, Smart Recommend, Feedback, and Changelog.
 
 **② Pick a model and launch your tool:**
@@ -148,17 +236,8 @@ If the active CLI tool is missing, FCM now catches it before launch, offers a ti
 ### Common scenarios
 
 ```bash
-# "I want the most reliable model right now"
-free-coding-models --fiable
-
-# "I want to configure Goose with an S-tier model"
-free-coding-models --goose --tier S
-
-# "I want NVIDIA's top models only"
-free-coding-models --origin nvidia --tier S
-
 # "I want the local web dashboard"
-free-coding-models --web
+free-coding-models --daemon
 
 # "I want one local endpoint that fails over between free models"
 free-coding-models --daemon-bg
@@ -174,7 +253,14 @@ free-coding-models --tier S --json | jq -r '.[0].modelId'
 free-coding-models --openclaw --origin groq
 ```
 
-When launching the web dashboard, `free-coding-models` prefers `http://localhost:3333`. If that port is already used by another app, it now auto-picks the next free local port and prints the exact URL to open.
+When launching the daemon (with `--daemon`), the web dashboard and router API are served from the same port. Configure tools with:
+
+| Field | Value |
+|-------|-------|
+| Router Base URL | `http://localhost:19280/v1` |
+| Dashboard URL | `http://localhost:19280/` |
+| Model | `fcm` |
+| API key | `fcm-local` |
 
 ### Smart Model Router
 
@@ -214,9 +300,20 @@ Router endpoints:
 | `GET /v1/models` | Return virtual models (`fcm`, `fcm:set-name`) |
 | `GET /health` | Daemon status JSON |
 | `GET /stats` | Routing, health, request log, and token stats |
-| `GET /stream/events` | Live SSE events for dashboard updates |
+| `GET /stream/events` | Live SSE events for router updates |
 | `POST /daemon/probe-mode` | Set probe mode with `{ "probeMode": "eco" | "balanced" | "aggressive" }` |
 
+**Web Dashboard endpoints** (served from the same port in `--daemon` mode):
+
+| Endpoint | Purpose |
+|----------|---------|
+| `GET /` | Web dashboard HTML |
+| `GET /api/models` | All model data with latency stats |
+| `GET /api/config` | Provider config (keys masked) |
+| `GET /api/events` | Live SSE events for dashboard |
+| `GET /api/key/:provider` | Reveal full API key for provider |
+| `POST /api/settings` | Save API keys and provider toggles |
+
 Routing behavior:
 
 - Priority order works immediately on cold start, then probes refine health scores over time.
@@ -505,6 +602,7 @@ Telemetry is enabled by default and can be disabled with any of the following:
     <td align="center" width="120"><a href="https://github.com/PhucTruong-ctrl"><img src="https://github.com/PhucTruong-ctrl.png?s=80" width="80" height="80" style="border-radius:50%" alt="PhucTruong-ctrl"></a></td>
     <td align="center" width="120"><a href="https://github.com/chindris-mihai-alexandru"><img src="https://avatars.githubusercontent.com/u/12643176?v=4&s=80" width="80" height="80" style="border-radius:50%" alt="chindris-mihai-alexandru"></a></td>
     <td align="center" width="120"><a href="https://github.com/serajbaltu"><img src="https://avatars.githubusercontent.com/u/90699173?v=4&s=80" width="80" height="80" style="border-radius:50%" alt="serajbaltu"></a></td>
+    <td align="center" width="120"><a href="https://github.com/stgreenb"><img src="https://avatars.githubusercontent.com/u/18483964?v=4&s=80" width="80" height="80" style="border-radius:50%" alt="stgreenb"></a></td>
   </tr>
   <tr>
     <td align="center"><a href="https://github.com/vava-nessa"><sub><b>vava-nessa</b></sub></a></td>
@@ -514,6 +612,7 @@ Telemetry is enabled by default and can be disabled with any of the following:
     <td align="center"><a href="https://github.com/PhucTruong-ctrl"><sub><b>PhucTruong-ctrl</b></sub></a></td>
     <td align="center"><a href="https://github.com/chindris-mihai-alexandru"><sub><b>chindris-mihai-alexandru</b></sub></a></td>
     <td align="center"><a href="https://github.com/serajbaltu"><sub><b>serajbaltu</b></sub></a></td>
+    <td align="center"><a href="https://github.com/stgreenb"><sub><b>stgreenb</b></sub></a></td>
   </tr>
 </table>
 

package/bin/free-coding-models.js

@@ -92,15 +92,7 @@ async function main() {
     process.exit(1);
   }
 
-  // 📖 --web mode: launch the web dashboard instead of the TUI
-  if (cliArgs.webMode) {
-    const { startWebServer } = await import('../web/server.js')
-    const port = parseInt(process.env.FCM_PORT || '3333', 10)
-    await startWebServer(port, { open: true })
-    return
-  }
-
-  // 📖 Load JSON config
+  // Load JSON config
   const config = loadConfig();
   ensureTelemetryConfig(config);
   ensureFavoritesConfig(config);

package/changelog/v0.1.1.md

@@ -0,0 +1,3 @@
+# Changelog v0.1.1 - 2026-02-21
+
+_No release notes available for this version._

package/changelog/v0.1.10.md

@@ -0,0 +1,3 @@
+# Changelog v0.1.10 - 2026-02-22
+
+_No release notes available for this version._

package/changelog/v0.1.11.md

@@ -0,0 +1,10 @@
+# Changelog v0.1.11 - 2026-02-22
+
+Add unit tests (59 tests), extract pure logic to lib/utils.js, fix GitHub Actions release workflow.
+
+- Extract getAvg, getVerdict, getUptime, sortResults, filterByTier, findBestModel, parseArgs to lib/utils.js
+- Add comprehensive test suite using node:test (sources integrity, core logic, CLI parsing, package sanity)
+- Fix release.yml: remove broken npm version patch, add version detection via git tags, add GitHub Release creation
+- Update AGENTS.md with test-first workflow instructions
+
+Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

package/changelog/v0.1.12.md

@@ -0,0 +1,16 @@
+# Changelog v0.1.12 - 2026-02-22
+
+### Added
+- Unit test suite: 59 tests across 11 suites using `node:test` (zero dependencies)
+- Tests cover: sources data integrity, core logic (getAvg, getVerdict, getUptime, filterByTier, sortResults, findBestModel), CLI arg parsing, package.json sanity
+- `lib/utils.js`: extracted pure logic functions from the monolithic CLI for testability
+- `pnpm test` script in package.json
+
+### Fixed
+- GitHub Actions release workflow: removed broken `npm version patch` loop, added version detection via git tags
+- GitHub Actions now creates a GitHub Release with auto-generated notes for each new version
+
+### Changed
+- AGENTS.md updated with test-first workflow: agents must run `pnpm test` before `pnpm start`
+
+---

package/changelog/v0.1.13.md

@@ -0,0 +1,3 @@
+# Changelog v0.1.13 - 2026-02-22
+
+_Intermediate patch release — part of the unit test suite addition (see [v0.1.12](../v0.1.12))._

package/changelog/v0.1.14.md

@@ -0,0 +1,3 @@
+# Changelog v0.1.14 - 2026-02-22
+
+_Intermediate patch release — part of the unit test suite addition (see [v0.1.12](../v0.1.12))._

package/changelog/v0.1.15.md

@@ -0,0 +1,3 @@
+# Changelog v0.1.15 - 2026-02-22
+
+_Intermediate patch release — part of the unit test suite addition (see [v0.1.12](../v0.1.12))._

package/changelog/v0.1.16.md

@@ -0,0 +1,13 @@
+# Changelog v0.1.16 - 2026-02-22
+
+### Added
+- OpenCode Desktop support: new `--opencode-desktop` flag and menu option to set model & open the Desktop app
+- "Read Changelogs" menu option when an update is available (opens GitHub releases page)
+- `startOpenCodeDesktop()` function — same config logic as CLI, launches via `open -a OpenCode`
+
+### Changed
+- Startup menu: "OpenCode" renamed to "OpenCode CLI", new "OpenCode Desktop" entry added
+- TUI mode badge: shows `[💻 CLI]` or `[🖥 Desktop]` or `[🦞 OpenClaw]`
+- Footer action hint adapts to desktop mode (`Enter→OpenDesktop`)
+
+---

package/changelog/v0.1.17.md

@@ -0,0 +1,3 @@
+# Changelog v0.1.17 - 2026-02-22
+
+_Intermediate patch release — part of the OpenCode Desktop iteration (see [v0.1.16](../v0.1.16))._

package/changelog/v0.1.18.md

@@ -0,0 +1,3 @@
+# Changelog v0.1.18 - 2026-02-22
+
+_Intermediate patch release — part of the OpenCode Desktop iteration (see [v0.1.16](../v0.1.16))._

package/changelog/v0.1.19.md

@@ -0,0 +1,3 @@
+# Changelog v0.1.19 - 2026-02-22
+
+_Intermediate patch release — part of the OpenCode Desktop iteration (see [v0.1.16](../v0.1.16))._

package/changelog/v0.1.2.md

@@ -0,0 +1,3 @@
+# Changelog v0.1.2 - 2026-02-22
+
+_No release notes available for this version._

package/changelog/v0.1.20.md

@@ -0,0 +1,14 @@
+# Changelog v0.1.20 - 2026-02-22
+
+### Added
+- **Dynamic tier filtering**: Use E/D keys to filter models by tier during runtime
+- Tier filter badge shown in header (e.g., `[Tier S]`)
+- E key elevates filter (show fewer, higher-tier models)
+- D key descends filter (show more, lower-tier models)
+- Preserves ping history when changing filters
+
+### Fixed
+- **Error 401 with --tier flag**: Fixed issue where using `--tier` alone would show selection menu instead of proceeding directly to TUI
+- Improved flag combination handling for better user experience
+
+---

package/changelog/v0.1.21.md

@@ -0,0 +1,12 @@
+# Changelog v0.1.21 - 2026-02-22
+
+### Refactored
+- **Simplified tier filtering architecture**: Replaced complex object recreation with simple `hidden` flag system
+- **Flags as shortcuts**: `--tier S` now just sets initial state instead of blocking dynamic filtering
+- **Dynamic filtering preserved**: E/D keys work seamlessly even when starting with `--tier` flag
+
+### Fixed
+- **Ping loop bug**: Fixed issue where filtered models weren't pinged due to using wrong results array
+- **Initial ping bug**: Fixed issue where initial ping used wrong results array
+
+---

package/changelog/v0.1.22.md

@@ -0,0 +1,6 @@
+# Changelog v0.1.22 - 2026-02-22
+
+### Changed
+- **Local changelogs**: "Read Changelogs" menu option now opens local `CHANGELOG.md` file instead of GitHub releases
+
+---

package/changelog/v0.1.23.md

@@ -0,0 +1,16 @@
+# Changelog v0.1.23 - 2026-02-22
+
+### Refactored
+- **Removed startup menu**: No more blocking mode selection menu at startup
+- **Default to OpenCode CLI**: App starts directly in CLI mode when no flags given
+- **Mode toggle in TUI**: Added Z key to cycle between CLI → Desktop → OpenClaw → CLI
+- **GitHub changelogs**: "Read Changelogs" option now opens GitHub URL instead of local file
+- **Auto-update by default**: When new version available without flags, auto-updates and relaunches
+- **Centered update menu**: Update notification appears only when needed, with clean centered layout
+
+### Changed
+- **Header display**: Shows `[💻 CLI] (Z to toggle)` with mode toggle hint
+- **Footer instructions**: Added "M Mode" to key bindings
+- **Update workflow**: Flags (`--opencode` etc.) still show update menu for compatibility
+
+---

package/changelog/v0.1.24.md

@@ -0,0 +1,9 @@
+# Changelog v0.1.24 - 2026-02-22
+
+## What's Changed
+* fix: TUI viewport scrolling for Ghostty and narrow terminals by @whit3rabbit in https://github.com/vava-nessa/free-coding-models/pull/10
+
+## New Contributors
+* @whit3rabbit made their first contribution in https://github.com/vava-nessa/free-coding-models/pull/10
+
+**Full Changelog**: https://github.com/vava-nessa/free-coding-models/compare/v0.1.23...v0.1.24

package/changelog/v0.1.25.md

@@ -0,0 +1,13 @@
+# Changelog v0.1.25 - 2026-02-22
+
+### Added
+- **Discord community link**: Added Discord invite to README and TUI footer
+- **Enhanced footer layout**: Improved footer with multi-line layout showing GitHub repo and Discord links
+- **Clickable author name**: "vava-nessa" is now clickable in terminal (opens GitHub profile)
+- **Release notes automation**: GitHub Actions now uses CHANGELOG.md content for release notes instead of auto-generated notes
+### Changed
+- **Tier filtering system**: Replaced E/D keys with T key that cycles through tier filters: all → S+/S → A+/A/A- → B+/B → C → all
+- **Footer text**: "Made with love by vava-nessa" with clickable links
+### Fixed
+- **Release workflow**: GitHub Releases now display proper changelog content instead of generic commit summaries
+---

package/changelog/v0.1.26.md

@@ -0,0 +1,6 @@
+# Changelog v0.1.26 - 2026-02-22
+
+### Changed
+- **Footer improvements**: Replaced "Repository GitHub" with "GitHub", "love" with 💖 emoji, and simplified Discord text
+- **README enhancement**: Added GitHub link section below Discord invite
+---

package/changelog/v0.1.27.md

@@ -0,0 +1,6 @@
+# Changelog v0.1.27 - 2026-02-22
+
+### Changed
+- **Footer redesign**: All links now on one line with clickable text: "Join our Discord!" and "Read the docs on GitHub"
+- **Improved UX**: Links use same clickable format as author name for consistent user experience
+---

package/changelog/v0.1.28.md

@@ -0,0 +1,5 @@
+# Changelog v0.1.28 - 2026-02-22
+
+### Added
+- **Footer emojis**: Added 💬 emoji before Discord link and ⭐ emoji before GitHub link for better visual appeal
+---

package/changelog/v0.1.29.md

@@ -0,0 +1,5 @@
+# Changelog v0.1.29 - 2026-02-22
+
+### Fixed
+- **Discord link correction**: Updated all Discord invite URLs to use permanent link https://discord.gg/WKA3TwYVuZ
+---

package/changelog/v0.1.3.md

@@ -0,0 +1,3 @@
+# Changelog v0.1.3 - 2026-02-22
+
+_No release notes available for this version._

package/changelog/v0.1.31.md

@@ -0,0 +1,10 @@
+# Changelog v0.1.31 - 2026-02-22
+
+### Added
+- **SWE-bench column**: Added new SWE-bench Verified score column showing coding performance for each model
+- **Color-coded column headers**: First letter of each column header is now colored (yellow) to indicate keyboard shortcut for sorting
+- **Keyboard shortcut improvements**: Changed P to L for latest ping sorting, added E for SWE-bench sorting
+### Changed
+- **Source name simplification**: Renamed "NVIDIA NIM" to "NIM" throughout the codebase
+- **Enhanced footer Discord link**: Discord link now displays in bright cyan color with "(link fixed)" indicator
+---

package/changelog/v0.1.32.md

@@ -0,0 +1,8 @@
+# Changelog v0.1.32 - 2026-02-22
+
+### Changed
+- **Column header improvements**: Fixed column alignment issues for better visual appearance
+- **Status renamed to Condition**: "Status" column renamed to "Condition" for clarity
+- **Keyboard shortcut updates**: S key now sorts by SWE-bench score, C key sorts by Condition
+- **Footer Discord text update**: Changed "Join our Discord!" to "Join Free-Coding-Models Discord!"
+---

package/changelog/v0.1.33.md

@@ -0,0 +1,5 @@
+# Changelog v0.1.33 - 2026-02-22
+
+### Fixed
+- **Column header alignment**: Fixed column headers misalignment issue caused by ANSI color codes interfering with text padding
+---

package/changelog/v0.1.34.md

@@ -0,0 +1,6 @@
+# Changelog v0.1.34 - 2026-02-22
+
+### Changed
+- **Condition renamed to Health**: Renamed "Condition" column to "Health" for better clarity
+- **Keyboard shortcut update**: H key now sorts by Health (instead of C for Condition)
+---

package/changelog/v0.1.35.md

@@ -0,0 +1,5 @@
+# Changelog v0.1.35 - 2026-02-22
+
+### Changed
+- **Column reorganization**: Reordered columns for better logical flow: Rank / Tier / SWE% / Model / Origin / Latest Ping / Avg Ping / Health / Verdict / Up%
+---

package/changelog/v0.1.36.md

@@ -0,0 +1,16 @@
+# Changelog v0.1.36 - 2026-02-22
+
+### Added
+- **SWE-bench Verified column**: Shows real SWE-bench Verified scores for all 44 models from official benchmarks
+- **Color-coded keyboard shortcuts**: First letter of each column header colored in yellow to indicate sorting key
+- **Heart and Coffee in footer**: "Made with 💖 & ☕ by vava-nessa"
+### Changed
+- **Column organization**: Reordered columns for better logical flow: Rank / Tier / SWE% / Model / Origin / Latest Ping / Avg Ping / Health / Verdict / Up%
+- **Health column**: Renamed from "Status" to "Health" with H key for sorting
+- **SWE-bench sorting**: S key now sorts by SWE-bench score
+- **Latest ping shortcut**: L key (instead of P) for sorting by latest ping
+- **Source name**: Simplified "NVIDIA NIM" to "NIM"
+### Fixed
+- **Column header alignment**: Fixed misalignment caused by ANSI color codes in headers
+- **Discord link**: Updated to permanent invite link https://discord.gg/WKA3TwYVuZ
+---

package/changelog/v0.1.37.md

@@ -0,0 +1,5 @@
+# Changelog v0.1.37 - 2026-02-22
+
+### Added
+- **Auto-update with sudo fallback**: When npm update fails due to permissions, automatically retries with sudo to complete the update
+---

package/changelog/v0.1.38.md

@@ -0,0 +1,9 @@
+# Changelog v0.1.38 - 2026-02-22
+
+### Fixed
+- **Cross-platform OpenCode integration**: Fixed OpenCode CLI and Desktop installation issues on Windows and Linux
+  - **Windows**: Fixed config path to use %APPDATA%\opencode\opencode.json with fallback to ~/.config
+  - **Linux**: Added support for snap, flatpak, and xdg-open to launch OpenCode Desktop
+  - **All platforms**: Properly detects OS and uses correct commands and paths
+  - **OpenCode Desktop**: Platform-specific launch commands (macOS: `open -a`, Windows: `start`, Linux: multiple methods)
+---

package/changelog/v0.1.39.md

@@ -0,0 +1,3 @@
+# Changelog v0.1.39 - 2026-02-22
+
+_Intermediate patch release — part of the cross-platform OpenCode fixes (see [v0.1.38](../v0.1.38))._

package/changelog/v0.1.4.md

@@ -0,0 +1,3 @@
+# Changelog v0.1.4 - 2026-02-22
+
+_No release notes available for this version._

package/changelog/v0.1.40.md

@@ -0,0 +1,3 @@
+# Changelog v0.1.40 - 2026-02-22
+
+_Intermediate patch release — part of the cross-platform OpenCode fixes (see [v0.1.38](../v0.1.38))._

package/changelog/v0.1.41.md

@@ -0,0 +1,11 @@
+# Changelog v0.1.41 - 2026-02-22
+
+### Changed
+- **sources.js data audit** — verified and corrected SWE-bench scores, tiers, and context windows across all NIM models:
+  - Devstral 2 123B: `S, 62.0%, 128k` → `S+, 72.2%, 256k` (official Mistral announcement)
+  - Mistral Large 675B: ctx `128k` → `256k`
+  - QwQ 32B: ctx `32k` → `131k`
+  - Llama 4 Maverick: ctx `128k` → `1M` (NVIDIA NIM confirmed)
+  - Llama 4 Scout: ctx `128k` → `10M` (NVIDIA NIM confirmed)
+  - GPT OSS 20B: ctx `32k` → `128k`
+---

package/changelog/v0.1.42.md

@@ -0,0 +1,3 @@
+# Changelog v0.1.42 - 2026-02-22
+
+_Intermediate patch release — part of the multi-provider iteration (see [v0.1.44](../v0.1.44) for full changelog)._

package/changelog/v0.1.43.md

@@ -0,0 +1,3 @@
+# Changelog v0.1.43 - 2026-02-22
+
+_Intermediate patch release — part of the multi-provider iteration (see [v0.1.44](../v0.1.44) for full changelog)._

package/changelog/v0.1.44.md

@@ -0,0 +1,17 @@
+# Changelog v0.1.44 - 2026-02-22
+
+### Added
+- **Multi-provider support** — Groq (6 models) and Cerebras (3 models) added alongside NVIDIA NIM, for 53 total models
+- **Multi-provider first-run wizard** — Steps through all 3 providers (NIM, Groq, Cerebras) on first launch; each is optional, Enter to skip; requires at least one key
+- **Settings screen (`P` key)** — New TUI overlay to manage API keys per provider, toggle providers on/off, and test keys with a live ping
+- **`lib/config.js`** — New JSON config system (`~/.free-coding-models.json`) replacing the old plain-text file
+  - Auto-migrates old `~/.free-coding-models` (plain nvidia key) on first run
+  - Stores keys per provider + per-provider enabled/disabled state
+  - `NVIDIA_API_KEY`, `GROQ_API_KEY`, `CEREBRAS_API_KEY` env vars override config
+- **Per-provider ping URLs** — `ping()` now accepts explicit endpoint URL; each provider has its own API endpoint in `sources.js`
+- **Provider name in Origin column** — Shows `NIM` / `Groq` / `Cerebras` instead of always `NIM`
+### Changed
+- `MODELS` flat array now includes `providerKey` as 6th element
+- State init filters models from disabled providers; rebuilds on settings close
+- Config file path changed from `~/.free-coding-models` to `~/.free-coding-models.json` (migration is automatic)
+---

package/changelog/v0.1.45.md

@@ -0,0 +1,6 @@
+# Changelog v0.1.45 - 2026-02-22
+
+### Fixed
+- **Discord notification**: Fixed GitHub Actions workflow crash (secrets context not allowed in step `if` conditions — now handled in the Node script directly)
+
+---

package/changelog/v0.1.46.md

@@ -0,0 +1,9 @@
+# Changelog v0.1.46 - 2026-02-23
+
+### Fixed
+- **Discord notification**: Fixed ECONNRESET error — drain response body with `res.resume()` and call `process.exit(0)` immediately after success so the Node process closes cleanly
+
+### Changed
+- **Discord link**: Updated invite URL to `https://discord.gg/5MbTnDC3Md` everywhere (README, TUI footer)
+
+---

package/changelog/v0.1.47.md

@@ -0,0 +1,6 @@
+# Changelog v0.1.47 - 2026-02-23
+
+### Changed
+- Release test
+
+---

package/changelog/v0.1.48.md

@@ -0,0 +1,11 @@
+# Changelog v0.1.48 - 2026-02-23
+
+### Fixed
+- **`--tier` CLI flag**: `parseArgs()` was never called in `main()`, so `--tier S` was silently ignored — now wired in and applied on TUI startup (thanks @whit3rabbit, PR #11)
+- **`--tier` value leaking into `apiKey`**: `parseArgs()` for-loop was capturing the tier value as the API key — fixed by skipping the value arg after `--tier`
+- **Ctrl+C not exiting**: sort key handler was intercepting all single-letter keypresses including ctrl-modified ones — added `!key.ctrl` guard so Ctrl+C reaches the exit handler (PR #11)
+
+### Added
+- Test verifying `--tier` value does not leak into `apiKey` (63 tests total)
+
+---

package/changelog/v0.1.49.md

@@ -0,0 +1,10 @@
+# Changelog v0.1.49 - 2026-02-23
+
+### Fixed
+- **Cerebras / Groq without API key**: models were being pinged with the fallback NVIDIA key, causing misleading `❌ 401` — now pings without auth header; 401 is treated as `🔑 NO KEY` (server reachable, latency shown dimly)
+- **Settings: entering an API key had no immediate effect**: after saving a key and closing Settings (Escape), models previously in `noauth` state are now immediately re-pinged with the new key
+
+### Changed
+- Ping without API key is now always attempted — a 401 response confirms the server is UP and shows real latency; `🔑 NO KEY` replaces the old `❌ 401` misleading error
+
+---

package/changelog/v0.1.5.md

@@ -0,0 +1,3 @@
+# Changelog v0.1.5 - 2026-02-22
+
+_No release notes available for this version._

package/changelog/v0.1.50.md

@@ -0,0 +1,7 @@
+# Changelog v0.1.50 - 2026-02-23
+
+### Fixed
+- **Groq/Cerebras models selected for OpenCode were launched as NVIDIA models**: `providerKey` was not passed in `userSelected` on Enter, causing all models to be prefixed with `nvidia/` regardless of their actual provider — now correctly uses `groq/model-id` and `cerebras/model-id`
+- **`startOpenCode` and `startOpenCodeDesktop`**: both functions now handle all 3 providers; Groq and Cerebras use OpenCode's built-in provider support (no custom config block needed, just `GROQ_API_KEY`/`CEREBRAS_API_KEY` env vars); NVIDIA retains its existing custom provider config flow
+
+---

package/changelog/v0.1.51.md

@@ -0,0 +1,4 @@
+# Changelog v0.1.51 - 2026-02-23
+
+### Fixed
+- **Groq/Cerebras models selected for OpenCode had no provider block**: even with the correct `groq/model-id` prefix, OpenCode couldn't use the model because no `provider.groq` block existed in `opencode.json` — now automatically creates the provider block (Groq: built-in with `apiKey: {env:GROQ_API_KEY}`; Cerebras: `@ai-sdk/openai-compatible` with baseURL) and registers the model in `provider.<key>.models`

package/changelog/v0.1.52.md

@@ -0,0 +1,10 @@
+# Changelog v0.1.52 - 2026-02-23
+
+### Fixed
+- **OpenCode model handoff** (PR #14 by @whit3rabbit): API keys from `~/.free-coding-models.json` were not passed to the OpenCode child process, causing silent fallback to the previous model. Also fixes Groq model ID mismatches (e.g. `kimi-k2-instruct` → `kimi-k2-instruct-0905`) via a new `OPENCODE_MODEL_MAP`
+- **OpenClaw nvidia provider missing models array** (PR #13 by @whit3rabbit): `startOpenClaw()` created the nvidia provider block without a `models` property, causing Zod schema validation to reject the config
+
+### Improved
+- **Discord link in TUI footer**: the invite URL is now displayed in plain text on a separate line so it's visible and copiable on terminals that don't support clickable links
+
+---