An Async Surface for Long-Running Agents
An async surface for long-running coding agents. Questions and blockers and reviews from every agent arrive in one inbox while they work. Answer and annotate there with Plannotator.
9 stories
An async surface for long-running coding agents. Questions and blockers and reviews from every agent arrive in one inbox while they work. Answer and annotate there with Plannotator.
Run /plannotator-last on a grill-me round: the questions open in your browser, you strike out assumptions and comment on fragments, and the feedback goes straight back to the agent.
Plannotator's /plannotator-annotate command now accepts URLs and HTML files, not just markdown. Fetch any web page, render local HTML, and annotate it with structured feedback for your coding agent.
Use your denial history to find feedback patterns, generate a personalized report, and automatically improve every future plan.
When your coding agent revises a plan, Plannotator now shows exactly what changed. Visual diffs, raw markdown diffs, and version history — across Claude Code, Codex, OpenCode, and Pi.
Plannotator now supports Pi, the minimal terminal coding agent. File-based plan mode, bash safety gating, progress tracking — all with the same browser review UI.
How Plannotator's /plannotator-review command gives you a browser-based diff viewer to annotate agent-generated code changes and send structured feedback back to the session.
How Plannotator's legacy URL sharing handles hash links, encrypted short links, and teammate annotations.
A plan review UI for Claude Code that lets you approve or request changes with annotated feedback.
Showing 9 of 9 stories
You're all caught up.
Get the top AI news and tools straight to your inbox every week.
Weekly · AI news
Join 15,000+ AI enthusiasts