OpenCodeReview combines deterministic checks and language model agents for code review with line-level comments.
Blinky
Blinky provides an open-source AI debugging agent integrated into Visual Studio Code.

About Blinky
Blinky is an open-source AI debugging agent that runs inside Visual Studio Code. It uses large language models (LLMs) to help identify and fix errors in backend code. The project is inspired by SWE-agent, and its core agent loop was adapted from that project. Blinky uses the VS Code API, the Language Server Protocol (LSP), and print statement debugging to locate and address bugs.
To use it, a developer installs the extension, describes the bug in the chat panel, and optionally lists repro steps. The agent then works on the code, and the developer can send feedback while it runs. When the agent finishes, the developer can accept or reject its changes. On first run, Blinky asks for an OpenAI API key, which is stored locally in the VS Code configuration. Requests are sent to OpenAI's servers, and the source does not describe other data handling.
Key features
- Uses LLMs to identify and attempt fixes for backend code errors
- Navigation tools such as GoToDefinition and GetAllReferences built on LSP
- GetFilesRelevantToEndpoint tool for navigating backend systems
- Verify tool runs user-specified repro steps until the test passes
- Accepts user feedback while the agent is running
- Match-and-replace file editing using line-numbered original text
- Open source, with the code available in a public GitHub repository
- Runs inside VS Code through a chat interface and the editor's own APIs
- Changes made by the agent are shown for review, with the option to accept or reject them
- Requires an OpenAI API key, and requests are sent to OpenAI's servers
- Support for other models such as LLaMA 3 and Claude is listed as a roadmap item, not a current feature
- Repro steps are optional but recommended, and the roadmap notes that writing them is tedious and time-consuming
Blinky at a glance
- Type
- Open-source project
- Pricing
- Not confirmed. Check the site
- Open source
- Yes · Source code
- Language
- English
- Last checked
- Oct 6, 2026 · Online when checked
- Listed
- Oct 5, 2026
Blinky alternatives
Tools and sites you could use instead, ranked by how closely they match.
This repository contains code for PR-Agent, an open source tool for reviewing pull requests.
ReviewCerberus analyzes Git branch differences and generates code review reports using AWS Bedrock or Anthropic API.
EvalView regression tests AI agents by snapshotting behavior and comparing tool calls in CI.
Greywall provides a deny-by-default sandbox for AI coding agents, with kernel-enforced isolation on Linux and macOS.
promptfoo tests prompts, agents, and retrieval systems with evaluations and security testing.
Popular in AI Code Review & Testing
Well-known listings from the same topic.
Blinky FAQ
How is Blinky related to SWE-agent?
What is the Verify tool?
How is the OpenAI API key handled?
Tell us and an editor will check it.
Claim the listing and suggest updates yourself. An editor checks every claim first.












