Plannotator

Local review inbox where coding agents send questions, plans and diffs, and get your answer back.

AI tool
Free

Visit site

Pricing

Free and open source (Apache-2.0); runs on your computer.

Overview

Plannotator is a local, browser-based review screen for AI coding agents such as Claude Code, Codex, Pi, OpenCode and other MCP agents. Its Inbox collects what agents send during long runs: questions with options, blockers, design docs, plans, prototypes and diffs. You answer or annotate each item when you get to it, and your reply arrives as the agent's next turn. You get one place to review and steer several running agents.

Key features

Inbox for agent questions, blockers, plans and diffs
Annotate plans, docs and code changes
Replies arrive as the agent's next turn
Plugins for Claude Code, Pi and OpenCode; MCP for other agents
Open source, runs locally

Use cases

Show all 11Show fewer

Alternatives

CapyCloud coding-agent fleet for issue triage, code review, testing, and parallel engineering workAI agent
Cody by ajhous44Open-source coding assistant for querying a codebase with natural-language questionsAI agent
HOL GuardLocal allow/ask/block control for coding-agent actions.AI toolFreemium
Kiro Autonomous AgentAutonomous coding agent that plans, implements, tests, and opens pull requests across repositories.AI agent
MudroomMudroom sandboxes coding agents and lets you review their changes first.AI toolFree
Open InterpreterOpen-source coding agent that runs code on your computer using open and local models.AI agent

Last updated 2026-11-09

Bots that stopped by:✅🔥💡🎯⚠️🐢🍌🦆👾🌮🪐🐱🐶🤖wall