Repo Prompt

About Repo Prompt

Repo Prompt is a macOS-native desktop application designed for developers who need reliable AI-assisted coding workflows tied to their local repositories. The tool provides a visual file tree with git-ignore filtering, instant previews, and regex search to select files, folders, or specific code sections. It includes an AI Context Builder that recommends relevant files, generates structured prompts and code maps, and estimates token usage to keep model context within limits. The Review and Apply mode converts model output into reviewable diffs, allowing developers to plan, inspect, and apply changes to disk in a controlled manner. Repo Prompt also integrates with an MCP server (supporting 15+ tools) so agents like Claude Desktop, Cursor, and Claude Code can read repository context, generate code maps, and apply changes programmatically. The application supports multiple model providers via user-supplied API keys and produces machine-readable, reversible patches that make large refactors or documentation updates easier to audit. Processing is kept local on macOS, ensuring data privacy and reducing dependency on cloud services.

Key features

• ✓Visual repo selection with git-ignore filtering
• ✓Instant file previews and regex search
• ✓AI Context Builder for structured prompts and code maps
• ✓Token usage estimates for model context limits
• ✓Review and Apply mode with diff-based change management
• ✓MCP server integration for programmatic agent access
• ✓Local macOS processing for data privacy
• ✓Support for multiple model providers via API keys
• ✓Machine-readable, reversible patch generation
• ✓Git-aware file selection and filtering

Use cases

• ✓Building structured AI prompts from local code repositories
• ✓Applying AI-generated code changes with diff-based review
• ✓Enabling AI coding agents to read repo context and apply changes

Frequently asked questions about Repo Prompt