lgtm is a TUI that runs beside an AI coding agent and shows its diffs live, letting developers review changes with vim-style motions and point at exact lines when replying. It is a standalone terminal tool, not LLM-powered itself; the LLM is the external coding agent whose output it reviews. It solves the problem of reviewing agent edits without leaving the editor workflow.
Visit project ↗A developer launched Revue, an open-source terminal-based code review tool that breaks large code changes into ordered, narrated chunks instead of a flat file list. It combines ideas from the tools hunk and stage-cli and lets any AI agent generate a guided tour via a skill, which users then step through in a TUI. It targets practitioners reviewing agent output without leaving their terminal multiplexer.
A developer launched Revue, an open-source terminal-based code review tool that breaks large code changes into ordered, narrated chunks instead of a flat file list. It combines ideas from the tools hunk and stage-cli and lets any AI agent generate a guided tour via a skill, which users then step through in a TUI. It targets practitioners reviewing agent output without leaving their terminal multiplexer.
Diffo opens a browser-based live diff of coding agent changes, connected to the agent session that wrote them. Developers comment on a line (e.g. "why did you touch the auth check?") and the agent answers in context or applies changes that update the diff. It solves the friction of switching back to chat to explain file references during code review.
Diffo opens a browser-based live diff of coding agent changes, connected to the agent session that wrote them. Developers comment on a line (e.g. "why did you touch the auth check?") and the agent answers in context or applies changes that update the diff. It solves the friction of switching back to chat to explain file references during code review.
A tool that helps developers using terminal-based AI coding agents (like Claude Code) keep track of changes made by the agent. It provides a file viewer with diffs to see what actually changed, and a view of the agent's actions, solving the problem of losing context when the agent iterates on code while the developer zones out.
A tool that helps developers using terminal-based AI coding agents (like Claude Code) keep track of changes made by the agent. It provides a file viewer with diffs to see what actually changed, and a view of the agent's actions, solving the problem of losing context when the agent iterates on code while the developer zones out.
Stvena is an open-source terminal tool that runs a live coding agent (Codex, Claude Code) alongside a Git review workspace. It lets developers read diffs, mark hunks reviewed, collect exact lines, and paste a single feedback draft back into the agent without switching to an IDE. It solves the slow review loop when iterating on multi-file agent changes from the terminal.
Stvena is an open-source terminal tool that runs a live coding agent (Codex, Claude Code) alongside a Git review workspace. It lets developers read diffs, mark hunks reviewed, collect exact lines, and paste a single feedback draft back into the agent without switching to an IDE. It solves the slow review loop when iterating on multi-file agent changes from the terminal.
The terminal tool reslop released version 0.1.x, implementing its Review phase: developers inspect AI-generated changes in an interactive terminal diff with intra-line highlighting, leave findings and todos, and prepare repair plans for an agent to execute. It targets uncommitted git changes, specific commits, or GitHub pull requests.
The terminal tool reslop released version 0.1.x, implementing its Review phase: developers inspect AI-generated changes in an interactive terminal diff with intra-line highlighting, leave findings and todos, and prepare repair plans for an agent to execute. It targets uncommitted git changes, specific commits, or GitHub pull requests.