Buildermark
UnclaimedMeasure and track the percentage of AI-generated code in your projects.
Visit WebsiteFreemiumVisit Website
TL;DR - Buildermark
- Quantifies AI-generated code by matching agent diffs with Git commits.
- Runs locally and is open source, ensuring data privacy and security.
- Tracks agent performance and provides attribution for AI-written code.
Pricing: Free plan available
Best for: Growing teams
Pros & Cons
Pros
- Ensures data privacy by running entirely locally
- Open-source and inspectable code
- Robust diff matching system handles formatting changes
- Supports multiple popular AI coding agents
- Provides valuable insights into AI agent effectiveness
Cons
- Team server for aggregation is not yet available
- Requires manual setup for importing from VMs and containers
Key Features
AI coding agent measurement and trackingConversation ratings for coding agentsAgent performance comparisonAutomatic import of chat history from coding agentsAutomatic import of Git commit historyFormatting-agnostic diff matching systemLocal-first operation with no cloud or accountsNative desktop applications (macOS, Windows, Linux)
Pricing
Freemium
Buildermark offers a generous free tier with optional paid upgrades for advanced features.
About Buildermark
By Toolradar Team·
Buildermark is a local-first, open-source tool designed to help developers and teams quantify the impact of AI coding agents on their codebase. It works by matching diffs from your coding agent conversations with your Git commit history to determine how much of your code is written by AI agents. This allows you to track the percentage of agent-written code over time and compare the performance of different agents across various projects.
The tool is built for privacy and data security, running entirely on your local machine without requiring cloud accounts or sending your data to external servers. It supports a range of popular coding agents like Claude Code, Codex, and Gemini, and offers features like conversation ratings and native notifications for immediate agent attribution. For teams, a self-hosted Team Server is planned to aggregate AI code metrics across an entire organization.
Reviews
No reviews yet. Be the first to review Buildermark!
Best Buildermark Alternatives
Top alternatives based on features, pricing, and user needs.
Explore More
Buildermark FAQ
How does Buildermark handle code formatting differences when matching agent diffs to commit diffs?
Buildermark utilizes a formatting-agnostic matching system. This means it's designed to be robust against common issues like auto-formatting and code reorganization, ensuring accurate attribution of agent-written code even if the formatting changes between the agent's output and the final commit.
What is the purpose of the 'Conversation ratings' feature, and how does it work?
The 'Conversation ratings' feature allows users to provide feedback on their interactions with coding agents. You can manually rate your coding agent conversations, or the agent itself can log feedback. This helps in comparing how different coding agents perform and understanding their effectiveness in various projects.
Since Buildermark runs locally, how will teams be able to aggregate AI code metrics across multiple developers?
For teams, a 'Team Server' is planned as a future paid, self-hosted solution. This server will allow organizations to aggregate Buildermark data from individual developers, providing a centralized view of AI code metrics across the entire team or organization.
Which specific coding agents are currently supported for importing chat history?
Buildermark currently supports importing chat history from Claude Code CLI, Claude Code Cloud, Codex CLI, Codex Cloud, Gemini CLI, and Cursor. Users can also file feature requests for additional agent support.
What is the technical architecture of Buildermark running on a local machine?
Buildermark operates as a local desktop application where an app container manages a Go server on localhost. This Go server then serves the web user interface, typically accessible at http://localhost:55022. This architecture ensures that all processing and data storage remain on the user's machine.
Source: buildermark.dev