Revdiff – TUI diff reviewer with inline annotations for AI agents
I built a terminal diff viewer for a workflow I couldn't do comfortably with existing tools: reviewing AI-generated code changes without leaving the terminal session where the agent runs, annotating what needs to change, and feeding those annotations straight back to the agent. Plenty of diff viewers exist, and some can even feed notes back to an agent, but none of them fit that flow for me - they pulled me out of the terminal into a separate app, or the round-trip back to the agent was clunky. Revdiff is how I solved it. From a Claude Code session I type `/revdiff main` or just say…
In plain words
Revdiff is a terminal-based diff viewer designed for reviewing AI-generated code changes without leaving the terminal session. It displays diffs as an overlay, allows users to add inline annotations on specific lines, hunks, or files, and sends those annotations directly back to AI agents for revision. The tool streamlines the feedback loop between code review and agent iteration, eliminating the need to switch between separate applications.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
I built a terminal diff viewer for a workflow I couldn't do comfortably with existing tools: reviewing AI-generated code changes without leaving the terminal session where the agent runs, annotating what needs to change, and feeding those annotations straight back to the agent. Plenty of diff viewers exist, and some can even feed notes back to an agent, but none of them fit that flow for me - they pulled me out of the terminal into a separate app, or the round-trip back to the agent was clunky. Revdiff is how I solved it. From a Claude Code session I type `/revdiff main` or just say "review diff for last 3 commits" - it opens as a terminal overlay on top of the agent session. I read the diff, drop annotations on any line, hunk, or file, then quit. The annotations go straight back to the agent, which picks them up and starts working on them. When the agent is done, revdiff reopens on the new changes so I can annotate again - and this loop continues until I quit without leaving any annotations. The same flow turned out to be useful for plans too. There's a separate `revdiff-planning` plugin that hooks into Claude Code's plan mode - the moment the agent finishes a plan and calls ExitPlanMode, revdiff opens automatically on the plan text. I annotate the parts I disagree with or want expanded, quit, and the agent revises the plan before writing any code. It's a two-pane TUI: file tree on the left, syntax-highlighted diff on the right. Vim-style navigation, /search, hunk jumping, word-level change highlighting, blame gutters, collapsed diff mode. Fully customizable - 7 bundled color themes, remappable keybindings, and per-color overrides via CLI flags, env vars, or config file. Works with git and Mercurial. It's a single binary - just `revdiff HEAD~3` to review your last 3 commits. Ready-to-use plugins are available for Claude Code (terminal overlay via tmux, Zellij, kitty, wezterm, Kaku, cmux, ghostty, iTerm2, or Emacs vterm - whichever you're in), Codex, and pi. Written in Go, open source (MIT). Install via brew: `brew install umputun/apps/revdiff` GitHub: https://github.com/umputun/revdiff Site: https://revdiff.com Would love to hear how others approach reviewing AI-generated diffs and PRs, any unusual use cases, and ideas for what to improve or add.
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 27d ago · cactuscompute.com
Launched alongside, April 2026
the whole month →
- AG
Thought the resources for GPU arch were lacking, so here we are
Life & fun · Apr 2026 · jaso1024.com
- IB
Built a ~9M param LLM from scratch to understand how they actually work. Vanilla transformer, 60K synthetic conversations, ~130 lines of PyTorch. Trains in 5 min on a free Colab T4. The fish thinks the meaning of life is food. Fork it and swap the personality for your own character.
AI · Apr 2026 · github.com
- BC
Life & fun · Apr 2026 · sam-burns.com
- IB
With social media and now AI, its important to keep the indie web alive. There are many people who write frequently. Blogosphere tries to highlight them by fetching the recent posts from personal blogs across many categories. There are two versions: Minimal (HN-inspired, fast, static): https://text.blogosphere.app/ Non-minimal: https://blogosphere.app/ If you don't find your blog (or your favorite ones), please add them. I will review and approve it.
AI · Apr 2026 · text.blogosphere.app