Add grammar-fixer extension

extensions/grammar-fixer/.eslintrc.json

@@ -0,0 +1,4 @@
+{
+  "root": true,
+  "extends": ["@raycast/eslint-config"]
+}

extensions/grammar-fixer/.gitignore

@@ -0,0 +1,3 @@
+node_modules/
+dist/
+.DS_Store

extensions/grammar-fixer/CHANGELOG.md

@@ -0,0 +1,13 @@
+# Grammar Fixer Changelog
+
+## [Initial Release] - {PR_MERGE_DATE}
+
+### Added
+- Fix grammar, spelling, and punctuation of selected text in-place via a single hotkey (`Cmd+Shift+G` on macOS, `Ctrl+Shift+G` on Windows)
+- Support for Anthropic, OpenAI, OpenRouter, Google, Groq, and Ollama providers
+- Raycast Pro support via `AI.ask()` — no API key needed
+- Custom model override via preferences
+- Extra instructions field for custom prompts (e.g. "Use British English")
+- Custom base URL support for OpenRouter and self-hosted endpoints
+- 5-minute response cache — repeated fixes on the same text are instant
+- Friendly error messages for common failures (invalid key, rate limit, quota exceeded, context too long, no text selected)

extensions/grammar-fixer/README.md

@@ -0,0 +1,61 @@
+<img src="assets/extension-icon.png" width="64" height="64" alt="Grammar Fixer" />
+
+# Grammar Fixer
+
+Fixing grammar in every message you send should not require switching to ChatGPT or Claude. Just select the text, hit `Cmd+Shift+G` (or `Ctrl+Shift+G` on Windows), and it's fixed in place — no copy, no paste, no context switch.
+
+## How It Works
+
+1. Select any text in any app
+2. Trigger **Grammar Fixer** with your hotkey
+3. The selected text is replaced with the corrected version instantly
+
+No UI. No interruptions. Just fixed text.
+
+## Recommended Hotkey Setup
+
+Raycast doesn't assign hotkeys automatically, so set it up once:
+
+1. Open Raycast and search **"Fix Grammar"**
+2. Hover the command → click `...` → **Add Hotkey**
+3. Press `Cmd+Shift+G` (macOS) or `Ctrl+Shift+G` (Windows)
+
+After that, the hotkey works system-wide from any app.
+
+## Setup
+
+Open Raycast preferences for Grammar Fixer and configure:
+
+| Preference | Description |
+|---|---|
+| **LLM Provider** | Choose from Anthropic, OpenAI, OpenRouter, Google, Groq, or Ollama |
+| **API Key** | Your API key for the selected provider |
+| **Model** | Model ID to use (leave blank for provider default) |
+| **Use Raycast AI** | Use Raycast's built-in AI if you have Raycast Pro (ignores provider/key) |
+| **Extra Instructions** | Optional instructions appended to the prompt (e.g. "Use British English") |
+| **Custom Base URL** | Override the API endpoint (useful for OpenRouter or self-hosted models) |
+
+## Supported Providers
+
+| Provider | Default Model |
+|---|---|
+| Anthropic | `claude-haiku-4-5-20251001` |
+| OpenAI | `gpt-4.1-nano` |
+| OpenRouter | `openai/gpt-4.1-nano` |
+| Google | `gemini-2.5-flash-lite` |
+| Groq | `llama-3.3-70b-versatile` |
+| Ollama (local) | `llama3.2` |
+
+You can override the model by typing any valid model ID in the **Model** preference field.
+
+## Raycast Pro
+
+If you have Raycast Pro, enable **Use Raycast AI** in preferences to use Raycast's built-in AI without needing an API key. You can still enter a placeholder value in the API Key field.
+
+## Ollama (Local Models)
+
+To use Ollama:
+1. Install [Ollama](https://ollama.com) and pull a model (e.g. `ollama pull llama3.2`)
+2. Select **Ollama (local)** as the provider
+3. Enter any value for the API Key
+4. Set the model name to match your pulled model