@@ -40,18 +40,17 @@ npx copilot-api --port 8080
4040
4141## Command Line Options
4242
43- The command accepts several command line options:
43+ The following command line options are available :
4444
45- | Option | Description | Default |
46- | ------------- | ------------------------------------ | ------- |
47- | --port, -p | Port to listen on | 4141 |
48- | --verbose, -v | Enable verbose logging | false |
49- | --log-file | File to log request/response details | - |
45+ | Option | Description | Default |
46+ | ------------- | ------------------------ | ------- |
47+ | --port, -p | Port to listen on | 4141 |
48+ | --verbose, -v | Enable verbose logging | false |
5049
51- Example with options :
50+ Example usage :
5251
5352``` sh
54- npx copilot-api@latest --port 8080 --verbose --log-file copilot-api.txt
53+ npx copilot-api@latest --port 8080 --verbose
5554```
5655
5756## Running from Source
@@ -70,15 +69,18 @@ bun run dev
7069bun run start
7170```
7271
73- ## Tips to not hit the rate limit
72+ ## Usage Tips
7473
75- - Use a free model from free provider like Gemini/Mistral/Openrouter for the weak model
76- - Rarely use architect mode
77- - Do not enable automatic yes in aider config
78- - Claude 3.7 thinking mode uses more tokens. Use it sparingly
74+ To avoid rate limiting and optimize your experience:
75+
76+ - Consider using free models (e.g., Gemini, Mistral, Openrouter) as the ` weak-model `
77+ - Use architect mode sparingly
78+ - Disable ` yes-always ` in your aider configuration
79+ - Be mindful that Claude 3.7 thinking mode consume more tokens
7980
8081## Roadmap
8182
82- - Manual approval for every request
83- - Rate limiting (only allow request every X seconds)
84- - Token counting
83+ - [ ] Manual request approval system
84+ - [ ] Rate limiting implementation
85+ - [ ] Token usage tracking and monitoring
86+ - [ ] Enhanced error handling and recovery
0 commit comments