Prompt CleanerCleans up your rambling prompts and sends them directly to Copilot Chat Installation Launch VS Code Quick Open ( Ctrl+P), paste the following command, and press enter.Copied to clipboard |
Prompt CleanerCleans up your messy, stream-of-consciousness prompts before sending them to Copilot Chat — removing filler words, fixing typos, trimming stray dots and self-corrections, then filling the Copilot Chat input with the cleaned version ready to send. Supports OpenAI, Anthropic Claude, xAI Grok, and any custom OpenAI-compatible endpoint (Ollama, LM Studio, OpenRouter, Hugging Face, and more). Commands
|
| Value | Description |
|---|---|
openai |
OpenAI GPT models (default) |
claude |
Anthropic Claude |
grok |
xAI Grok |
custom |
Any OpenAI-compatible endpoint — local or remote |
Each provider has its own settings group. You can configure all of them and just toggle provider to switch — no need to re-enter keys.
OpenAI
| Setting | Default | Description |
|---|---|---|
copilotPromptCleaner.openai.apiKey |
(empty) | API key from platform.openai.com |
copilotPromptCleaner.openai.model |
gpt-4o |
Model name (e.g. gpt-4o, gpt-4o-mini, gpt-4-turbo) |
"copilotPromptCleaner.provider": "openai",
"copilotPromptCleaner.openai.apiKey": "sk-...",
"copilotPromptCleaner.openai.model": "gpt-4o"
Anthropic Claude
| Setting | Default | Description |
|---|---|---|
copilotPromptCleaner.claude.apiKey |
(empty) | API key from console.anthropic.com |
copilotPromptCleaner.claude.model |
claude-haiku-4-5 |
Model name (e.g. claude-haiku-4-5, claude-sonnet-4-5, claude-opus-4-5) |
"copilotPromptCleaner.provider": "claude",
"copilotPromptCleaner.claude.apiKey": "sk-ant-...",
"copilotPromptCleaner.claude.model": "claude-haiku-4-5"
xAI Grok
| Setting | Default | Description |
|---|---|---|
copilotPromptCleaner.grok.apiKey |
(empty) | API key from console.x.ai |
copilotPromptCleaner.grok.model |
grok-4.3 |
Model name (e.g. grok-4.3, grok-3) |
"copilotPromptCleaner.provider": "grok",
"copilotPromptCleaner.grok.apiKey": "xai-...",
"copilotPromptCleaner.grok.model": "grok-4.3"
Custom endpoint
Use this for any OpenAI-compatible server — local or remote. Both endpoint and model are required. apiKey is optional.
| Setting | Default | Description |
|---|---|---|
copilotPromptCleaner.custom.endpoint |
(empty) | Base URL without /v1 |
copilotPromptCleaner.custom.model |
(empty) | Exact model name as required by the provider |
copilotPromptCleaner.custom.apiKey |
(empty) | API key — leave empty for servers that don't require auth |
Ollama (local, no key needed)
"copilotPromptCleaner.provider": "custom",
"copilotPromptCleaner.custom.endpoint": "http://localhost:11434",
"copilotPromptCleaner.custom.model": "llama3"
LM Studio (local)
"copilotPromptCleaner.provider": "custom",
"copilotPromptCleaner.custom.endpoint": "http://localhost:1234",
"copilotPromptCleaner.custom.model": "phi3"
OpenRouter
"copilotPromptCleaner.provider": "custom",
"copilotPromptCleaner.custom.endpoint": "https://openrouter.ai/api",
"copilotPromptCleaner.custom.model": "mistralai/mixtral-8x7b-instruct",
"copilotPromptCleaner.custom.apiKey": "sk-or-..."
Hugging Face
"copilotPromptCleaner.provider": "custom",
"copilotPromptCleaner.custom.endpoint": "https://api-inference.huggingface.co",
"copilotPromptCleaner.custom.model": "HuggingFaceH4/zephyr-7b-beta",
"copilotPromptCleaner.custom.apiKey": "hf_..."
Tip: Store API keys in User Settings (
Cmd+Shift+P→ Open User Settings JSON), not workspace settings, so they are never committed to git.