Plannotator v0.28.51

Plannotator is an open-source platform designed to review and validate plans and code generated by AI coding agents.

The platform provides a centralized interface where developers can:

• Review AI-generated plans

• Annotate proposed actions

• Approve or reject tasks

• Inspect code changes

• Provide structured feedback

• Improve agent outputs

• Maintain human oversight

Rather than replacing developers, Plannotator aims to keep humans involved in critical decision-making during AI-assisted development.

As AI coding agents become increasingly capable, developers face a growing challenge: ensuring that AI-generated plans and code changes align with project requirements before they are executed. Plannotator addresses this problem by acting as a review layer between developers and AI coding agents.

Instead of generating code itself, Plannotator focuses on making AI-driven development more transparent and controllable. It allows teams to review, annotate, approve, reject, and refine agent-generated plans and code changes before they affect a project.

Download Plannotator v0.28.51 - Software Mirrors
Plannotator v0.28.51 for Windows plannotator-win32-x64.exe | 202.27 MB plannotator-win32-arm64.exe | 187.91 MB plannotator-paste-win32-x64.exe | 109.94 MB plannotator-paste-win32-arm64.exe | 95.58 MB
Plannotator v0.28.51 for macOS plannotator-paste-darwin-x64 | 93.62 MB plannotator-paste-darwin-arm64 | 85.03 MB plannotator-darwin-x64 | 185.95 MB plannotator-darwin-arm64 | 178.08 MB
Plannotator v0.28.51 for Linux plannotator-paste-linux-x64 | 104.07 MB plannotator-paste-linux-arm64 | 104.17 MB plannotator-linux-x64 | 196.4 MB plannotator-linux-arm64 | 196.49 MB
Plannotator v0.28.51 Source Code Plannotator v0.28.51 Source code (zip) Plannotator v0.28.51 Source code (tar.gz)
Plannotator v0.28.51 Release Notes: Follow @plannotator on X for updates Missed recent releases? v0.28.50: A faster, lighter Inbox, session names across the Inbox, initiatives on the Board, earlier comments on a file, global comments on Guided Reviews v0.28.49: A checklist for each thread, a pill that shows who gets your reply, a calmer full screen, Inbox replies in T3 Code, --title for review and annotate v0.28.48: Reply attachments that follow the draft, links to any web page on Board cards, instant theme switches, a daily anonymous ping, a faster Inbox v0.28.47: Board lanes, Settings → Rows, a new thread header, a "Make it yours" tour step, Request guided review, Jira links on Board cards v0.28.46: Snooze a thread, Settings → Accounts, long pastes become file tiles, HTML pages on a phone, hide the repository owner v0.28.45: Plannotator Modern theme, a thread panel view, session names in the To picker, thread links open in place, comment on diagram parts v0.28.44: The iPhone keyboard no longer moves the Inbox, pictures stay clear of the option card, What's new over your tailnet v0.28.43: Voice dictation, combine and rename projects, Mark unread, full-screen threads, Q for open questions, search by type v0.28.42: Choose how your Inbox list looks, contents for long documents, link the Inboxes on two computers, clearer search, documents outside the project open from your own machine v0.28.40: Update works with update checks off, the thread bar shows the thread's name, agents before the project in list rows v0.28.39: Reply from full screen, thread names lead the Inbox list, one project per Claude Code session, a redesigned code review sidebar, Call flow as cards v0.28.38: Attach files to a reply in the Inbox, swipe through a question's picture choices, easier-to-read list rows, code on hover in Call Flow and Semantic Changes What's New in v0.28.51 Starred threads, a note when an agent's session has ended, "Back from snooze", an Ask box in Settings → Decision data, and expand arrows in guided reviews. 25 PRs, all from @backnotprop. Plannotator Inbox Starred threads Pins are now stars. A filled star replaces the pin, and the menus say Star thread and Unstar thread. Click a row's star (tap it on a phone) to unstar the thread. This doesn't open the thread or mark it read, and the toast offers Undo, or press Z. The second click of a double click does nothing, so it can't open the thread you just unstarred. You still star a thread from its right-click menu or its ⋯ menu, and your existing pins show as stars. On a touch screen the star's target is 45 px wide. A tap near the end of a row's preview may unstar the thread instead of opening it, and Undo covers that. #2685 See when an agent's session has ended Before, nothing told you that the session behind a thread had ended, and replies to it just waited. Now the row says Session ended. The thread's note says what to do, with Send to another session and Archive beside Reply. A reply already saved for that session says so: "Saved for Pi. That session has ended; it gets this only if it's resumed." Send to another session moves the saved reply, and any earlier ones waiting with it, to a live session. That session then takes over the thread, so your next Send goes there. The Inbox checks that the session really ended before it moves anything. Agents that connect only through MCP don't report when they end, so their rows show nothing. #2637 Back from snooze When a snooze runs out, the open thread says "Back from snooze · 8:00 AM". If there is news, the line sits on the divider above the first new message. If nothing is new, it sits above the last message. The thread opens at that line, and the line stays until you leave the thread or snooze it again. It doesn't show after an Unsnooze or a reply. Also, an archived thread no longer says "Snoozed until …" after that time has passed. #2675 Ask about a past decision Settings → Decision data has an Ask box. Type a question about something you decided earlier and press Ask. The question goes to a running agent as a new message, with instructions to search the threads your Inbox has stored, in every project. The agent answers with what was asked, the choices, what you chose and when, your own words and a link to the thread. If the answer isn't stored, the agent is told to say so and not to guess. The answer comes back as an ordinary thread, and the Inbox opens it. If several sessions are running, pick one from the list. If none is running, the box says "No agent is running" and offers Copy. The Inbox stores and shows only your words, not the instructions. #2681 Show more of a file in a guided review A guided review in a thread now has the "N unmodified lines" rows and arrows from Plannotator's code review, so you can open the lines around a change, up to the whole file. Each file's box is as tall as its content, with no height limit, and the page scrolls. Nothing is read until you press an arrow, and the file comes from your own repository, not from GitHub. If your repository doesn't have that version of the file, the arrows go away on the first press. You can comment on an opened line, and your agent receives the comment marked "(not in the diff)" with the code quoted. A large file no longer freezes the page when it appears, because it is no longer read whole and highlighted on the page's main thread. In headless Chromium, with other work running on the machine, a 5,000-line changed file showed its first row in 422 ms instead of 6,826 ms, and the longest freeze went from 5,080 ms to 61 ms. #2635 "These might help" in the Feedback dialog As you type in the Feedback box, the dialog lists related links under "These might help": Plannotator features with a one-line description, open issues, pull requests and release note headings. It shows five at most, or three on a phone. The search runs in the window, so nothing you type reaches a server before you send. If nothing is similar, the form looks as before. Words you typed but didn't send come back when you reopen the dialog, until the window closes. #2641 Feedback sends the names of the agents you use Privacy change: feedback you send from the Inbox's Feedback dialog now also includes the names of the agents you use with the Inbox, for example "Claude Code, Codex". Only the names are sent. No session ids, paths, thread titles, counts or text go with them. The Inbox builds the list from the agent messages it has stored and uses Plannotator's own name for each agent. It never sends an agent it doesn't know or a name an agent gave itself. The list holds at most 8 names, most recently used first, and the page can't add names of its own. The dialog's "what is sent" line now says so: "Sends your message (and the category and email, if you add them), Plannotator {version}, the name of your OS and the names of the agents you use. Nothing from your threads or files." plannotator feedback on the command line is unchanged and sends no names. #2712 Smaller changes On a phone, the floating box over a file in full screen keeps its foot on one row. The open-question count, the comment pill and "Open in thread" get shorter, in that order, before anything wraps. Its buttons are now 44 px targets. If dictation is still being transcribed when the box closes, the words now land in that thread's reply instead of being lost. #2636 While dictation shows, keys you type in the reply box, New message or full screen's card are no longer lost, and undo works under the highlight on new words. #2673 When an agent recommends a choice by its leading letter or number, such as Recommended: A for A · Yes, the Inbox now marks that choice as recommended instead of drawing an extra option. #2724 The Drafts page header now just says Drafts. #2670 Attachments now also refuse the NTFS spellings of a .env file, such as .env::$DATA. #2665 Not yet on This release also carries work on starting agents from the Inbox, code review in the Inbox and Shape boards. All three stay off by default, and their switches stay hidden until a later release. #2451, #2678, #2683, #2640, #2661, #2665, #2679, #2660, #2710, #2715 Plannotator Long command output arrives whole When their output went to a pipe or a socket, plannotator card list (text and --json), plannotator inbox projects --json and plannotator guide export --out - could lose the end of their output and still exit 0. An agent reading a long card list could be missing cards without knowing it. These commands now exit only after all of their output is written, and the text is the same. One difference: if the program reading the output stops reading but keeps the pipe open, the command now waits until it reads or closes, as cat does. #2653 Smaller changes Internal: test runs no longer read the real home folder or skill folders, and a slow CI setup step now has a time limit and retries. Nothing changes for users. #2680, #2686, #2687 Install / Update macOS / Linux: bash curl -fsSL https://plannotator.ai/install.sh | bash Windows: powershell irm https://plannotator.ai/install.ps1 | iex Claude Code Plugin: The install script above also updates the plugin when claude is on your PATH. Otherwise, in a terminal: bash claude plugin marketplace update plannotator claude plugin update plannotator@plannotator Then restart Claude Code. Inside Claude Code, run /plugin marketplace update plannotator, then open /plugin → Installed → plannotator → Update now. Pi: bash pi update npm:@plannotator/pi-extension OpenCode: Re-run the install script above. It also clears the OpenCode 2 plugin cache. What's Changed feat(inbox): These might help, under the Feedback box by @backnotprop in #2641 Shape: the whiteboard, and What to know replaces Requirements (off by default) by @backnotprop in #2660 Inbox: Trust folder button for Claude Code launches (off by default) by @backnotprop in #2451 Wait for the reader before exiting: card, inbox projects, guide export, shape kit, help by @backnotprop in #2653 fix(inbox): the Drafts header says only Drafts by @backnotprop in #2670 Keep the Inbox's floating box in one tidy row on a phone by @backnotprop in #2636 Inbox: Open in for any file a reference names (off by default) by @backnotprop in #2665 Inbox Reviews: rename and route (#reviews, #pulls alias, &new, off line) (off by default) by @backnotprop in #2661 Inbox voice: keep keys typed while dictation shows, and undo under its highlight by @backnotprop in #2673 Keep test runs out of the real skill folders by @backnotprop in #2680 ci(inbox-e2e): limit and retry the browsers' system libraries step by @backnotprop in #2687 fix(inbox): keep a PR review's unsent comments and Viewed until a new commit (off by default) by @backnotprop in #2640 feat(inbox): the guided review's rail moves into its top bar (off by default) by @backnotprop in #2679 fix(inbox): say why a folder whose trust would cover home can't be trusted (off by default) by @backnotprop in #2683 Inbox: say "Back from snooze" in the open thread by @backnotprop in #2675 Inbox: say when a thread's session has ended, and let the reply go elsewhere by @backnotprop in #2637 Tests: give every Inbox a test starts a scratch HOME by @backnotprop in #2686 Inbox launch: an answer on the card no longer flashes the Terminal line (off by default) by @backnotprop in #2678 Feedback from the Inbox also sends the names of the agents you use by @backnotprop in #2712 Ask from Settings -> Decision data by @backnotprop in #2681 Shape: icons, logos and drawn things by name, and the whiteboard words (off by default) by @backnotprop in #2715 Expand arrows on the Inbox's PR page and guided reviews by @backnotprop in #2635 Name a question choice by its leading letter before a middle dot by @backnotprop in #2724 feat(inbox): threads are starred, and a click on the star unstars by @backnotprop in #2685 Shape: a crosshair for comments, lazy drags, and Add to reply (off by default) by @backnotprop in #2710 Community @UberMouse asked for taller code blocks in guided reviews in #2643. #2635 partly answers it: in the Inbox's guided reviews, each file is now as tall as its content. The issue stays open for the rest. Full Changelog: v0.28.50...v0.28.51

Key Features of Plannotator

Plan Review System

One of Plannotator's core capabilities is reviewing plans generated by AI agents before execution.

Developers can examine:

• Proposed tasks

• Implementation strategies

• Agent reasoning

• Planned file modifications

• Workflow sequences

This visibility helps reduce unintended changes and costly mistakes.

Annotation Tools

The platform allows users to add comments, notes, and guidance directly to AI-generated plans.

These annotations can be used to:

• Clarify requirements

• Correct misunderstandings

• Provide context

• Guide future agent actions

AI Code Review

Plannotator extends the review process beyond planning by supporting inspection of generated code.

Developers can:

• Review modifications

• Analyze diffs

• Leave comments

• Request revisions

• Validate implementation details

This workflow resembles modern pull-request review systems.

Human-in-the-Loop Workflows

A major design goal is ensuring that AI actions remain subject to human approval.

Organizations can establish review processes where important actions require validation before execution.

Open Source Foundation

Plannotator is open source, allowing teams to inspect, modify, and self-host the platform according to their needs.

This transparency is particularly valuable for organizations adopting AI-assisted software development.

User Experience

The interface is designed around review workflows rather than direct code generation.

Instead of interacting with a chatbot, users primarily:

1. Receive agent-generated plans

2. Review proposed actions

3. Add feedback

4. Approve or reject changes

5. Monitor execution results

The workflow feels familiar to developers accustomed to pull requests, code reviews, and project planning tools.

Productivity Benefits

As AI coding tools become more autonomous, review processes become increasingly important.

Plannotator helps organizations:

• Reduce risky AI actions

• Improve code quality

• Increase accountability

• Preserve architectural consistency

• Encourage collaboration between developers and AI agents

For teams adopting AI-driven development, these safeguards can be as valuable as the coding agents themselves.

Collaboration Features

The platform supports collaborative review workflows where multiple team members can participate in evaluating AI-generated outputs.

This allows:

• Peer review

• Team approval processes

• Shared annotations

• Collective decision-making

Such features are especially useful for larger engineering teams.

Performance

Because Plannotator focuses on workflow management and review rather than model inference, performance largely depends on the connected AI agents and integrations.

The platform itself is lightweight and primarily serves as an orchestration and review layer.

Open Source Advantages

Being open source provides several benefits:

• Transparent development

• Self-hosting capabilities

• Custom integrations

• Community contributions

• Vendor independence

Organizations concerned about compliance, security, or proprietary workflows may find these advantages particularly appealing.

Limitations

Plannotator is designed as a companion tool rather than a complete AI development platform.

Common limitations include:

• Requires external AI coding agents

• Best suited for teams already using AI-assisted development

• Smaller ecosystem than mature developer platforms

• Additional review steps may slow rapid prototyping

• Some users may prefer fully autonomous workflows

The software delivers the most value in environments where oversight and quality control are priorities.

Pros

• Improves transparency of AI-generated plans

• Supports structured review workflows

• Human-in-the-loop design

• Useful annotation system

• Open source

• Self-hosting support

• Familiar review experience for developers

• Helps reduce AI-generated mistakes

Cons

• Not a standalone coding agent

• Requires integration with AI development tools

• Smaller community than established developer platforms

• Adds review overhead to workflows

• Best suited for teams rather than casual users

Who Should Use Plannotator?

Plannotator is ideal for:

• Software development teams

• Engineering managers

• AI-assisted development workflows

• Organizations adopting coding agents

• Open-source projects

• Teams prioritizing code quality and governance

It is particularly valuable for environments where AI-generated code requires oversight before reaching production systems.

Plannotator fills an increasingly important role in the AI development ecosystem by providing visibility and control over AI-generated plans and code changes. Its focus on human oversight, structured reviews, and collaborative workflows makes it a useful companion for modern coding agents. While it is not a replacement for AI coding tools themselves, it offers a practical solution for teams seeking greater confidence and accountability in AI-assisted software development.

Free

Download

Software Informations:
Developer:
backnotprop
Operating System:
Windows / macOS / Linux
Date Added:
2026-11-06T23:07:18.196Z
Categories:

Related to Plannotator v0.28.51:

Post a Comment/Report Broken Link: