Plannotator v0.27.13 Download

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 v0.27.13 - Software Mirrors
Plannotator v0.27.13 for Windows plannotator-paste-win32-arm64.exe \| 90.2 MB plannotator-paste-win32-x64.exe \| 93.92 MB plannotator-win32-arm64.exe \| 143.62 MB plannotator-win32-x64.exe \| 147.33 MB
Plannotator v0.27.13 for macOS plannotator-darwin-arm64 \| 114.35 MB plannotator-darwin-x64 \| 119.39 MB plannotator-paste-darwin-arm64 \| 60.51 MB plannotator-paste-darwin-x64 \| 65.97 MB
Plannotator v0.27.13 for Linux plannotator-linux-arm64 \| 142.73 MB plannotator-linux-x64 \| 143.62 MB plannotator-paste-linux-arm64 \| 89.35 MB plannotator-paste-linux-x64 \| 90.21 MB
Plannotator v0.27.13 for Other plannotator-0.27.13-release-sbom.cdx.json \| 1.61 MB
Plannotator v0.27.13 Release Notes: Follow @plannotator on X for updates Missed recent releases? v0.27.12: Unified decision control, token hover cards, local-vs-remote diff, approval notes v0.27.11: OpenCode server leak fix, durable local feedback archive, unknown-subcommand fix v0.27.10: Auto-viewed files on scroll, annotation undo/redo, OpenCode 2 slash commands restored, npm 12 agent terminal fix v0.27.9: WebMCP browser-agent tools, HTML refresh from disk, host seams, lazy renderers, Windows uninstall fix v0.27.8: Pi keeps its prompt cache across plan transitions, thumbs-up returns to HTML annotation, embed picker seam v0.27.7: Pi host crash fix on Windows, Call Flow tree cap, jj fork-point base, plannotator knowledge skill + llms.txt v0.27.6: Live app annotation lands on Pi, one interaction model for HTML pages v0.27.5: Annotate your running app, Agent TUI placement, collapsed lockfiles, VS Code theme fix v0.27.4: Portable Guided Review exports, guides.show share links, guide CLI, jj Call Flow v0.27.3: Folder watcher freeze fix on large repos, first SBOM-attested release pipeline v0.27.2: Mobile plan and code review, Codex CLI 0.147 fix, configurable markdown extensions v0.27.1: Open-in-editor launch fix, file headers respect Viewed/Git-add visibility toggles What's New in v0.27.13 Six pull requests, four of them from community contributors, two from first-time contributors. The headline answers a request straight from X: open a code review against the base you actually mean, not always the trunk. Alongside it: a security hardening of the document endpoint, the end of our longest-standing CI flake, and a Amp delivery fix. Open a review on a specific base `plannotator review` always opened against the detected trunk, even when you were reviewing one layer of a stacked branch. A user on X put it plainly: "it always opens vs. main even if it explicitly understands that I'm reviewing a stack." Now the review can open exactly where you point it: `bash stack: main → feature/part-1 → feature/part-2 (you are here) plannotator review --base feature/part-1 opens on "All changes since feature/part-1": just what this layer adds plannotator review --base feature/part-1 --diff-type merge-base the committed-only view of the same layer` `--base` takes anything git resolves: a branch, `origin/branch`, a tag, a commit, `HEAD~3`. `--diff-type` picks the opening view from the same nine modes the dropdown offers. Both are a starting state, nothing more: the session opens there, everything stays changeable in the UI, and neither flag ever touches your saved defaults. A base that does not exist fails at launch with a clear message and a "did you mean" suggestion instead of silently producing a wrong diff. This matters most when an agent opens the review for you. The agent knows which layer it just built, so it can hand you a review already looking at the right thing instead of telling you which dropdown to click. The installed skills teach exactly that: reviewing a stack layer, pass `--base` . The flags work across every host that launches reviews through the CLI, and error honestly on surfaces where a base has no meaning (jj, GitButler, Perforce, multi-repo workspaces, and PR URLs, whose base comes from the platform). #1484 One behavior change shipped with this. The review command used to silently ignore flags it did not recognize; a typo like `--bse main` opened a review as if you had typed nothing, and in the worst case an unknown flag could swallow the PR URL next to it. Unknown dash-prefixed flags now fail loudly with a usage hint, matching how annotate has always behaved. Plain words stay tolerated, so slash-command hosts that forward your raw sentence keep working. OpenCode and Pi slash commands also now reject CLI-only transport flags such as `--tailscale` that they previously accepted and ignored. #1483 Symlinks can no longer read outside the project A symlink committed into a repository you are reviewing could point anywhere on your disk, and the endpoint that serves linked documents would follow it: the requested path looked like it was inside the project, but the content came from outside it. Document reads are now gated on real-path containment in both server runtimes: a path must land inside the project both as written and after resolving symlinks, or the request is refused. Legitimate symlink setups keep working, including symlinked project roots (macOS temp directories) and links that resolve within the project. Three edge behaviors changed on purpose: a symlink escaping the project returns 403 on every path through the endpoint, HTML files over the 2MB annotate cap now return 413 where one path previously served them in full, and a file that exists but cannot be opened returns 500 instead of 404. Contributed by @bendrucker in a first contribution that went well beyond its own scope: along the way he reproduced our longest-standing CI flake, disproved our working theory with actual experiments, and traced it to the real root cause below. #1437 The CI flake is dead, and tests stay out of your data Four annotate-server tests had been failing intermittently on CI for weeks (#1464), passing on rerun, resisting diagnosis. The root cause turned out to be a single line: shared storage captured the data directory once at module import, so a test that imported it under a temporary override poisoned every later test in the process, and test-file discovery order decided who got hit. The same freeze meant a full test run could write into a contributor's real `~/.plannotator`. Storage now resolves its directory per call, and the test suite sandboxes `PLANNOTATOR_DATA_DIR` for every run, with regression tests locking both halves down. Two contributors converged on the same root cause independently within a day, from different starting points: @FNDEVVE working from our test-isolation issue, and @bendrucker debugging his own PR's CI failures. Closes #1455 and #1464. #1473 Amp: feedback is routed by decision, never guessed from prose The Amp plugin classified review outcomes by searching the rendered feedback text for phrases, so a genuine comment like "this path has no feedback loop, add one" pattern-matched as no-action and was silently dropped. The CLI now offers `plannotator review --json`, emitting one structured `{ decision, message }` record from the same builder as the plaintext output, and Amp routes purely on the decision field. The prose classifier is gone. An outdated CLI produces a recoverable update notice with the captured output, never a guessed decision. Contributed by @FNDEVVE, closing #1456. #1476 Additional Changes Selection toolbar stays on screen on phones: selecting short text near a screen edge in compact touch layouts overflowed the floating toolbar past the viewport, clipping its buttons. It now clamps inside the screen (respecting notch safe areas) and keeps its centered position whenever that fits. Desktop placement is untouched. Contributed by @katya4oyu in their first contribution. #1471 The view dropdown follows the live base: the "All changes since " option label was baked at session start from the detected trunk and never updated, so it could contradict the base picker sitting next to it. It now renders from the active base. 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: Run `/plugin` in Claude Code, find plannotator, and click "Update now". Pi: Update `@plannotator/pi-extension` to 0.27.13 and restart Pi. OpenCode: Clear cache and restart: `bash rm -rf ~/.bun/install/cache/@plannotator` What's Changed feat(review): open a review on a specific base and diff type in #1484 fix(review): strict argument parsing for unknown review flags in #1483 fix(doc): deny symlink escapes on /api/doc reads by @bendrucker in #1437 fix(test): isolate test-run data from contributor history by @FNDEVVE in #1473 fix(amp): relay structured decisions without prose classification by @FNDEVVE in #1476 fix(ui): keep selection toolbar within compact touch viewport by @katya4oyu in #1471 New Contributors @bendrucker made their first contribution in #1437 @katya4oyu made their first contribution in #1471 Community This release is mostly the community's. @bendrucker shipped the symlink containment work through two rebases we caused, and when CI failed on his branch he did the diagnosis himself: reproduced it deterministically, tested and disproved our working theory, and root-caused a flake that had dogged this project for weeks, then filed the follow-up (#1477) for the remaining cases. @FNDEVVE swept our own issue tracker and closed two filed issues in as many days, with fixes careful enough that one of them independently matched bendrucker's root cause line for line. @katya4oyu fixed a real phone papercut with a minimal, well-tested change that read the repo's conventions closely enough to regenerate a build manifest most first PRs miss. And the headline feature exists because a user on X told us the review always opened against the wrong base when reviewing a stack. They were right. Thank you. Plannotator gets better because you tell us where it falls short. Full Changelog: v0.27.12...v0.27.13

Download Plannotator v0.27.13 - Software Mirrors

Plannotator v0.27.13 Release Notes:

Follow @plannotator on X for updates

Missed recent releases?

v0.27.12: Unified decision control, token hover cards, local-vs-remote diff, approval notes

v0.27.11: OpenCode server leak fix, durable local feedback archive, unknown-subcommand fix

v0.27.10: Auto-viewed files on scroll, annotation undo/redo, OpenCode 2 slash commands restored, npm 12 agent terminal fix

v0.27.9: WebMCP browser-agent tools, HTML refresh from disk, host seams, lazy renderers, Windows uninstall fix

v0.27.8: Pi keeps its prompt cache across plan transitions, thumbs-up returns to HTML annotation, embed picker seam

v0.27.7: Pi host crash fix on Windows, Call Flow tree cap, jj fork-point base, plannotator knowledge skill + llms.txt

v0.27.6: Live app annotation lands on Pi, one interaction model for HTML pages

v0.27.5: Annotate your running app, Agent TUI placement, collapsed lockfiles, VS Code theme fix

v0.27.4: Portable Guided Review exports, guides.show share links, guide CLI, jj Call Flow

v0.27.3: Folder watcher freeze fix on large repos, first SBOM-attested release pipeline

v0.27.2: Mobile plan and code review, Codex CLI 0.147 fix, configurable markdown extensions

v0.27.1: Open-in-editor launch fix, file headers respect Viewed/Git-add visibility toggles

What's New in v0.27.13

Six pull requests, four of them from community contributors, two from first-time contributors. The headline answers a request straight from X: open a code review against the base you actually mean, not always the trunk. Alongside it: a security hardening of the document endpoint, the end of our longest-standing CI flake, and a Amp delivery fix.

Open a review on a specific base

plannotator review always opened against the detected trunk, even when you were reviewing one layer of a stacked branch. A user on X put it plainly: "it always opens vs. main even if it explicitly understands that I'm reviewing a stack." Now the review can open exactly where you point it:

bash
stack: main → feature/part-1 → feature/part-2 (you are here)
plannotator review --base feature/part-1
opens on "All changes since feature/part-1": just what this layer adds
plannotator review --base feature/part-1 --diff-type merge-base
the committed-only view of the same layer

--base takes anything git resolves: a branch, origin/branch, a tag, a commit, HEAD~3. --diff-type picks the opening view from the same nine modes the dropdown offers. Both are a starting state, nothing more: the session opens there, everything stays changeable in the UI, and neither flag ever touches your saved defaults. A base that does not exist fails at launch with a clear message and a "did you mean" suggestion instead of silently producing a wrong diff. This matters most when an agent opens the review for you. The agent knows which layer it just built, so it can hand you a review already looking at the right thing instead of telling you which dropdown to click. The installed skills teach exactly that: reviewing a stack layer, pass --base . The flags work across every host that launches reviews through the CLI, and error honestly on surfaces where a base has no meaning (jj, GitButler, Perforce, multi-repo workspaces, and PR URLs, whose base comes from the platform). #1484 One behavior change shipped with this. The review command used to silently ignore flags it did not recognize; a typo like --bse main opened a review as if you had typed nothing, and in the worst case an unknown flag could swallow the PR URL next to it. Unknown dash-prefixed flags now fail loudly with a usage hint, matching how annotate has always behaved. Plain words stay tolerated, so slash-command hosts that forward your raw sentence keep working. OpenCode and Pi slash commands also now reject CLI-only transport flags such as --tailscale that they previously accepted and ignored. #1483

Symlinks can no longer read outside the project

A symlink committed into a repository you are reviewing could point anywhere on your disk, and the endpoint that serves linked documents would follow it: the requested path looked like it was inside the project, but the content came from outside it. Document reads are now gated on real-path containment in both server runtimes: a path must land inside the project both as written and after resolving symlinks, or the request is refused. Legitimate symlink setups keep working, including symlinked project roots (macOS temp directories) and links that resolve within the project. Three edge behaviors changed on purpose: a symlink escaping the project returns 403 on every path through the endpoint, HTML files over the 2MB annotate cap now return 413 where one path previously served them in full, and a file that exists but cannot be opened returns 500 instead of 404. Contributed by @bendrucker in a first contribution that went well beyond its own scope: along the way he reproduced our longest-standing CI flake, disproved our working theory with actual experiments, and traced it to the real root cause below. #1437

The CI flake is dead, and tests stay out of your data

Four annotate-server tests had been failing intermittently on CI for weeks (#1464), passing on rerun, resisting diagnosis. The root cause turned out to be a single line: shared storage captured the data directory once at module import, so a test that imported it under a temporary override poisoned every later test in the process, and test-file discovery order decided who got hit. The same freeze meant a full test run could write into a contributor's real ~/.plannotator. Storage now resolves its directory per call, and the test suite sandboxes PLANNOTATOR_DATA_DIR for every run, with regression tests locking both halves down. Two contributors converged on the same root cause independently within a day, from different starting points: @FNDEVVE working from our test-isolation issue, and @bendrucker debugging his own PR's CI failures. Closes #1455 and #1464. #1473

Amp: feedback is routed by decision, never guessed from prose

The Amp plugin classified review outcomes by searching the rendered feedback text for phrases, so a genuine comment like "this path has no feedback loop, add one" pattern-matched as no-action and was silently dropped. The CLI now offers plannotator review --json, emitting one structured { decision, message } record from the same builder as the plaintext output, and Amp routes purely on the decision field. The prose classifier is gone. An outdated CLI produces a recoverable update notice with the captured output, never a guessed decision. Contributed by @FNDEVVE, closing #1456. #1476

Additional Changes

Selection toolbar stays on screen on phones: selecting short text near a screen edge in compact touch layouts overflowed the floating toolbar past the viewport, clipping its buttons. It now clamps inside the screen (respecting notch safe areas) and keeps its centered position whenever that fits. Desktop placement is untouched. Contributed by @katya4oyu in their first contribution. #1471

The view dropdown follows the live base: the "All changes since " option label was baked at session start from the detected trunk and never updated, so it could contradict the base picker sitting next to it. It now renders from the active base.

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: Run /plugin in Claude Code, find plannotator, and click "Update now". Pi: Update @plannotator/pi-extension to 0.27.13 and restart Pi. OpenCode: Clear cache and restart:

bash
rm -rf ~/.bun/install/cache/@plannotator

What's Changed

feat(review): open a review on a specific base and diff type in #1484

fix(review): strict argument parsing for unknown review flags in #1483

fix(doc): deny symlink escapes on /api/doc reads by @bendrucker in #1437

fix(test): isolate test-run data from contributor history by @FNDEVVE in #1473

fix(amp): relay structured decisions without prose classification by @FNDEVVE in #1476

fix(ui): keep selection toolbar within compact touch viewport by @katya4oyu in #1471

New Contributors

@bendrucker made their first contribution in #1437

@katya4oyu made their first contribution in #1471

Community

This release is mostly the community's. @bendrucker shipped the symlink containment work through two rebases we caused, and when CI failed on his branch he did the diagnosis himself: reproduced it deterministically, tested and disproved our working theory, and root-caused a flake that had dogged this project for weeks, then filed the follow-up (#1477) for the remaining cases. @FNDEVVE swept our own issue tracker and closed two filed issues in as many days, with fixes careful enough that one of them independently matched bendrucker's root cause line for line. @katya4oyu fixed a real phone papercut with a minimal, well-tested change that read the repo's conventions closely enough to regenerate a build manifest most first PRs miss. And the headline feature exists because a user on X told us the review always opened against the wrong base when reviewing a stack. They were right. Thank you. Plannotator gets better because you tell us where it falls short. Full Changelog: v0.27.12...v0.27.13

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:

Receive agent-generated plans
Review proposed actions
Add feedback
Approve or reject changes
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:
2026-09-10T23:00:23.627Z
Categories:
Developer Tools

Software Mirrors - Blog

Download Plannotator v0.27.13 - Software Mirrors

Plannotator v0.27.13 for Windows

Plannotator v0.27.13 for macOS

Plannotator v0.27.13 for Linux

Plannotator v0.27.13 for Other