Coding & DevelopmentāØgeminiš¤claudeāØļøcursoršuniversal
CLI Yes Automation Skill
A wrapper tool that automates interactions with the Claude CLI by automatically handling common prompts and responses
Quick Start
Install this Skill
npx skills add snomiao/agent-yes# CLI Yes Automation Skill
A skill for automating AI CLI tool interactions by handling common prompts and managing continuous operation.
## Description
This skill helps you work with the `agent-yes` wrapper tool, which automates interactions with various AI CLI tools (Claude, Gemini, Codex, Copilot, Cursor, Grok, Qwen) by automatically responding to common prompts and keeping the tools running continuously.
## When to Use This Skill
- Setting up automated AI CLI workflows
- Configuring continuous operation with AI assistants
- Implementing auto-response patterns for yes/no prompts
- Managing crash recovery and idle detection
- Running AI CLI tools in automation scripts or CI/CD pipelines
## Key Capabilities
### Multi-CLI Support
Works with multiple AI coding assistants:
- Claude Code (Anthropic) - Industry-leading performance
- Gemini CLI (Google) - Free tier with generous limits
- Codex CLI (OpenAI/Microsoft) - Cloud-based collaboration
- Copilot CLI (GitHub) - Seamless GitHub integration
- Cursor CLI - Multi-model support with RAG
- Grok CLI (xAI) - Real-time data access
- Qwen Code CLI (Alibaba) - Open source, high performance
### Automation Features
- **Auto-Response**: Automatically responds "Yes" to common prompts
- **Continuous Operation**: Keeps AI running until task completion
- **Crash Recovery**: Automatic process restart on crashes
- **Idle Detection**: Optional auto-exit when AI becomes idle
- **Interactive Control**: Queue prompts or cancel with ESC/Ctrl+C
## Usage Examples
### Basic Command Line Usage
```bash
# Use Claude (default)
agent-yes claude -- run all tests and commit current changes
bunx agent-yes claude "Solve TODO.md"
# Use other AI tools
agent-yes codex -- refactor this function
agent-yes grok -- help me with this code
agent-yes copilot -- generate unit tests
agent-yes cursor -- optimize performance
agent-yes gemini -- debug this code
agent-yes qwen -- implement new feature
# Auto-exit when idle (for automation)
agent-yes claude --exit-on-idle=60s "run all tests and commit current changes"
```
### Library Usage in Node.js
```typescript
import cliYes from "agent-yes";
// Use Claude
await cliYes({
prompt: "help me solve all todos in my codebase",
cli: "claude",
cliArgs: ["--verbose"],
exitOnIdle: 30000, // exit after 30 seconds of idle
continueOnCrash: true,
logFile: "claude.log",
});
// Use other tools
await cliYes({
prompt: "debug this function",
cli: "gemini",
exitOnIdle: 60000,
});
```
## Configuration Options
- `--cli=<tool>`: Specify AI CLI tool (claude, gemini, codex, copilot, cursor, grok, qwen)
- `--exit-on-idle=<duration>`: Auto-exit after specified idle time (e.g., "60s", "5m")
- Custom CLI args can be passed through for tool-specific options
## Security Considerations
ā ļø **Important**: Only run on trusted repositories. This tool:
- Automatically responds to prompts without user confirmation
- Can execute commands automatically
- May be vulnerable to prompt injection attacks in malicious code/files
Always review repositories before running automated tools.
## Implementation Details
Uses `node-pty` or `bun-pty` to manage AI CLI processes with:
- **Pattern matching**: Detects ready states, prompts, and errors
- **Auto-response system**: Sends "Yes" to common prompts
- **Process lifecycle management**: Handles crashes and graceful exits
- **Tool-specific configurations**: Custom patterns for each CLI
## Installation
```bash
# Install the wrapper tool globally
npm install agent-yes -g
# Install your preferred AI CLI
npm install -g @anthropic-ai/claude-code # Claude
npm install -g @vibe-kit/grok-cli # Grok
# See documentation for other CLI installation
```
## Best Practices
1. **Start small**: Test with simple tasks before complex automation
2. **Use idle timeout**: Set `--exit-on-idle` for automated scripts
3. **Review output**: Check logs and results regularly
4. **Trust repositories only**: Never run on untrusted code
5. **Choose the right CLI**: Match tool to task requirements
- Complex tasks ā Claude Code
- Budget-conscious ā Gemini or Qwen
- GitHub integration ā Copilot
- Team collaboration ā Codex or Cursor
## Resources
- GitHub: https://github.com/snomiao/claude-yes
- Claude Code: https://www.anthropic.com/claude-code
- Issue Tracker: https://github.com/snomiao/claude-yes/issues
## License
MIT - See project repository for details
Repository
9stars
5forks
Details
Related Skills
More coding & development skills you might like
Reticle
Coding & DevelopmentAI agents can generate code, but still struggle to understand what they build. Reticle brings Jev-style machine-native runtime perception to web & desktop applications.
š¤claudešuniversal
1.2k
190
View
agentscope-java
Coding & DevelopmentExpert Java developer skill for AgentScope Java framework - a reactive, message-driven multi-agent system built on Project Reactor. Use when working with reactive programming, LLM integration, agent orchestration, multi-agent systems, or when the user mentions AgentScope, ReActAgent, Mono/Flux, Project Reactor, or Java agent development. Specializes in non-blocking code, tool integration, hooks, pipelines, and production-ready agent applications.
šuniversal
6.0k
1.5k
View
ospec
Coding & DevelopmentDocument-driven OSpec workflow for initialization, change/goal routing, validation, archiving, and durable project knowledge.
āØgeminiš¤claudeš¬gptšuniversal
524
34
View
misakanet-failure-memory
Coding & Developmentš A zero-dependency, git-backed micro-lesson library for AI Agents to asynchronously share and search verified debugging experience. Python stdlib only. | https://misakanet.org
š¤claudeāØļøcursoršuniversal
424
158
View