Ad-Hoc Prompts

Use prompt to run arbitrary prompts with AI agents for tasks beyond code review:

roborev prompt "Explain the architecture of this codebase"
roborev prompt --wait "What does the main function do?"
roborev prompt --agent claude-code "Refactor error handling in main.go"
roborev prompt --reasoning thorough "Find potential security issues"
roborev prompt --agentic "Add error handling to main.go"
cat instructions.txt | roborev prompt --wait

Flags

Flag	Description
--wait	Wait for job to complete and show result
--agent	Agent to use (default: from config)
--reasoning	Reasoning level: fast, standard, or thorough
--no-context	Don’t include repository context in prompt
--agentic	Enable agentic mode (allow file edits and commands)
--yolo	Alias for --agentic
--quiet	Suppress output (just enqueue)

Review vs Agentic Mode

By default, prompts run in review mode (read-only). Use --agentic (or --yolo) to enable write operations (file edits, bash commands) for tasks that need to modify your codebase.

Mode	Tools Available	Use Case
Review (default)	Read, Glob, Grep	Analysis, explanations, finding issues
Agentic	Read, Glob, Grep, Edit, Write, Bash	Refactoring, implementing features, fixes

Repository Context

By default, prompts include context about the repository:

• Repository name and path
• Any project guidelines from .roborev.toml

Use --no-context for raw prompts without this context.

Examples

Ask Questions

roborev prompt "What design patterns are used in this codebase?"
roborev prompt --wait "Explain the authentication flow"

Code Analysis

roborev prompt --reasoning thorough "Find potential security issues"
roborev prompt "What are the main entry points of this application?"

Make Changes (Agentic)

roborev prompt --agentic "Add input validation to the user controller"
roborev prompt --yolo "Refactor the database layer to use connection pooling"

Piped Input

echo "Add comprehensive error handling" | roborev prompt --agentic --wait
cat todo.txt | roborev prompt --wait

See Also

• Agent Modes - Details on review vs agentic modes
• Auto-Fixing - Automated issue resolution