Reado

A calm place to read code.

reado.watermelon-studio.it

_{The loop: leave a comment on a line → Send review → the agent resolves it and marks the task done.}

Most IDEs are built for writing code. Reado inverts that: the primary experience is reading, and the primary action is leaving durable comments anchored to precise points in the code. Those comments are not throwaway notes — they are persistent, lifecycle-bearing artifacts that an AI agent (Claude Code, Codex or Copilot) resolves, and that accumulate into a consultable knowledge base.

The mental model is an inverted code review: you are the reviewer (read, annotate), the AI is the committer (implements the fixes).

Reado is built with Tauri 2, React 19 and a Rust backend. It runs on macOS, Linux and Windows.

Download

Grab the latest signed build for your platform from the Releases page — macOS (.dmg, signed & notarized), Linux (.AppImage / .deb / .rpm) and Windows (.exe / .msi). Reado updates itself from signed releases after that.

Status

Reado is in active development. The current focus is the read → annotate → AI-resolve loop. See openspec/changes/add-reado-mvp for the full specification and the sequenced task list.

What works today:

• Open any local folder (git repository or not) from a recent-projects launcher.
• Gitignore-aware file tree with a "show hidden" toggle.
• A read-first CodeMirror 6 viewer with broad syntax highlighting, large-file virtualization, a line-wrap toggle, comfortable reading width, and a soft landing highlight when you jump to a location.
• Dedicated rendering for markdown, images and foldable JSON.
• Fuzzy file open (Cmd/Ctrl+P), full-text project search (ripgrep when present, with a built-in pure-Rust fallback), and a command palette (Cmd/Ctrl+K).
• Code reading aids: an outline of file symbols, sticky scope headers, go-to-definition (Cmd/Ctrl+click / F12), go-to-line, and Format Document through the project's own formatters (Biome/Prettier/rustfmt/…).
• Git introspection with stage/unstage/discard/commit, inline blame, and a local diff view.
• Four research-grounded themes (dark, light, high-contrast, sepia) with manual, follow-system, and time-of-day selection.
• Italian and English UI.
• Anchored comments as an external overlay: leave a comment on a line range (hover the + or press the shortcut), with types, states, task/note, threads, and a gutter marker. Comments are stored as .md files under .reado/ and survive external edits (git-diff-free fuzzy re-anchoring; orphans never point at the wrong line).
• An integrated terminal (real PTYs, multiple tabs) with one-click launch of claude, codex and copilot, and Send review to hand your open tasks to the agent.
• A knowledge base unifying the project's docs, specs (OpenSpec / speckit), and the notes captured while reading, with full-text search and a knowledge graph linking comments, files, specs and docs.
• An MCP server and a reado CLI — how agents reach Reado. The MCP server (auto-wired into each agent) exposes your tasks, comments, reading progress and bookmarks as read-only context, plus browser_* tools to inspect and drive the in-app browser preview; the CLI is how the agent acts — resolving tasks, consulting the knowledge base, proposing review artifacts. Both are the same reado binary, which must be on the agent's PATH.

The AI loop

Reado's core loop is read → annotate → AI-resolve:

1. You read code and leave comments. Comments flagged as tasks are the work list; notes stay out of the agent's way.
2. Open the terminal (Cmd/Ctrl+J), launch Claude, Codex or Copilot, then click Send review. Reado injects a prompt pointing the agent at your open tasks.
3. The agent reads your tasks and comments — from the reado://tasks and reado://comments MCP resources, or reado task list — makes the changes, and marks each done with reado task done <id> (or reado task fail <id> "<reason>"). Reado's watcher reflects the result live, and resolved comments move to history.

The reado binary is the stable contract — the on-disk format can evolve without breaking the agents. It serves both the MCP server (reado mcp, auto-wired into each agent's config on project open) and the CLI the agent calls, so it must be on the agent's PATH for the AI loop to work. The packaged app bundles it: install from Settings → Command-line tool (links reado into ~/.local/bin, VS Code style). From a source checkout, build and link it directly:

scripts/install-cli.sh           # builds release + links into ~/.local/bin
reado --help

Actions (CLI): reado task list|show|done|fail|link, reado comment add|reply|search, and reado kb list|show|search (to consult the docs and specs before resolving). Context (MCP): the reado://tasks, reado://comments, reado://reading-progress and reado://bookmarks resources, plus browser_* tools for the in-app preview. Agent identity comes from $READO_AGENT (Reado sets it when launching an agent).

An agent plugin in plugin/ teaches Claude Code (and Codex, via AGENTS.md) this contract so the agent resolves tasks correctly. See plugin/README.md to install it. Other agents (e.g. Copilot) still get the contract from the Send review prompt Reado injects, as long as the reado CLI is installed.

Keyboard shortcuts

Shortcut	Action
Cmd/Ctrl + P	Go to file (fuzzy)
Cmd/Ctrl + K	Command palette
Cmd/Ctrl + Shift+F	Full-text project search
Cmd/Ctrl + ,	Settings

Development

Prerequisites: Node.js 22+, pnpm, the Rust toolchain, and the Tauri system dependencies for your OS. ripgrep (rg) on your PATH makes full-text search faster; without it Reado falls back to a built-in scanner.

pnpm install        # install frontend dependencies
pnpm tauri dev      # run the app in development
pnpm tauri build    # produce a native installer for the current platform

Useful checks:

pnpm typecheck                                   # TypeScript
cargo test    --manifest-path src-tauri/Cargo.toml
cargo clippy  --manifest-path src-tauri/Cargo.toml --all-targets -- -D warnings

Contributing

Reado aims to be a friendly open-source project. Contributions are welcome — see CONTRIBUTING.md and our Code of Conduct. Security issues: see SECURITY.md. All code, comments and documentation are written in English.

License

MIT © Reado contributors