[{"data":1,"prerenderedAt":178},["ShallowReactive",2],{"site-nav-marketing":3,"feature-review":36,"features-all-nav":161},{"features":4,"compare":23},[5,8,11,14,17,20],{"label":6,"to":7},"Sessions & status","\u002Ffeatures\u002Fsessions-and-status",{"label":9,"to":10},"Review","\u002Ffeatures\u002Freview",{"label":12,"to":13},"Worktrees & hooks","\u002Ffeatures\u002Fworktrees-and-hooks",{"label":15,"to":16},"Git & issues","\u002Ffeatures\u002Fgit-and-issues",{"label":18,"to":19},"Stacks","\u002Ffeatures\u002Fstacks",{"label":21,"to":22},"Teams","\u002Ffeatures\u002Fteams",[24,27,30,33],{"label":25,"to":26},"vs Claude Code","\u002Fcompare\u002Fclaude-code",{"label":28,"to":29},"vs Claude Squad and Nimbalyst","\u002Fcompare\u002Fclaude-squad-and-nimbalyst",{"label":31,"to":32},"vs Conductor","\u002Fcompare\u002Fconductor",{"label":34,"to":35},"vs Cursor","\u002Fcompare\u002Fcursor",{"id":37,"title":38,"body":39,"description":78,"draft":79,"extension":80,"faq":81,"hero":94,"meta":97,"navTitle":9,"navigation":98,"order":75,"path":10,"sections":99,"seo":158,"slug":108,"stem":159,"__hash__":160},"features\u002Ffeatures\u002Freview.md","Review AI agent changes like a pull request",{"type":40,"value":41,"toc":73},"minimark",[42,47],[43,44,46],"h2",{"id":45},"read-more","Read more",[48,49,50,59,66],"ul",{},[51,52,53,58],"li",{},[54,55,57],"a",{"href":56},"\u002Fdocs\u002Freview","Review & diffs",": the full Review tab, from the diff to the review-and-fix cycle",[51,60,61,65],{},[54,62,64],{"href":63},"\u002Fdocs\u002Fpull-requests","Pull requests",": opening pull requests and tracking their status",[51,67,68,72],{},[54,69,71],{"href":70},"\u002Fdocs\u002Fsettings","Settings",": review instructions, models and the round limit",{"title":74,"searchDepth":75,"depth":75,"links":76},"",2,[77],{"id":45,"depth":75,"text":46},"A PR-style diff of each Claude Code session's worktree, inline comments the agent acts on, AI review, a review-and-fix cycle and the PR conversation.",false,"md",[82,85,88,91],{"question":83,"answer":84},"Do my review notes leave my machine?","No. Notes are stored locally and exist so the agent can read them. A note reaches GitHub or Bitbucket only when you publish it to the pull request, after a confirmation that shows the text and the repository.",{"question":86,"answer":87},"Does the review-and-fix cycle push my code?","No. Plexus never pushes for you. The cycle edits files in the session's worktree, and you read the result before it goes anywhere.",{"question":89,"answer":90},"What runs the AI review?","A headless Claude Code instance on your machine, using your own Claude Code login. You choose the model and effort level for the review and for the fix agent in Settings → Git.",{"question":92,"answer":93},"Which pull requests can I review in Plexus?","Pull requests on GitHub and Bitbucket, opened from a branch in the repository itself. Pull requests from forks are not supported yet. Bitbucket has no review verdicts, so its review comments arrive as ordinary threads.",{"eyebrow":9,"headline":95,"lead":96},"Review the agent's work before anyone else sees it.","The Review tab diffs a session's worktree against its base branch, like a pull request. Comment on lines and send the comments to the agent, run an AI review, or let a review-and-fix cycle run on its own.",{},true,[100,112,123,131,139,150],{"id":101,"heading":102,"body":103,"state":104,"docsPath":105,"sim":106},"diff","A pull request view of the worktree","The Review tab compares the session's worktree with its base, using the remote-tracking branch when there is one, so it matches what the pull request will show. Untracked files are included. Switch between unified and split views; in both, the words that changed inside a line are marked. The file list shows each file's change type, line counts and how many comments are open or resolved on it.","neutral","\u002Fdocs\u002Freview#the-diff-view",{"view":107,"tab":108,"anchors":109},"sessions","review",[110,111,101],"review-bar","review-files",{"id":113,"heading":114,"body":115,"state":116,"docsPath":117,"sim":118},"comments","Comments the agent reads","Hover a line, write a note and add it. Notes stay on your machine. When you are done, Send N comments to agent turns every open note into one prompt and writes it to the running agent, which picks it up on its next turn. Sent notes are marked resolved, so the file list keeps a record of what you handed over.","working","\u002Fdocs\u002Freview#inline-comments",{"view":107,"tab":108,"anchors":119},[120,121,122],"diff-thread","thread-send","send-review-comments",{"id":124,"heading":125,"body":126,"state":127,"docsPath":128,"sim":129},"ai-review","AI review on demand","Run AI review hands the diff to a separate, headless Claude, which looks for bugs, security issues and risky changes and leaves its findings as inline comments. You write the review instructions, globally or per workspace or project. Plexus supplies the output format, so the comments land on the right lines whatever your instructions say.","delegating","\u002Fdocs\u002Freview#ai-review",{"view":107,"tab":108,"anchors":130},[124],{"id":132,"heading":133,"body":134,"state":127,"docsPath":135,"sim":136},"cycle","Review and fix, unattended","The review-and-fix cycle runs an AI review, starts a separate headless agent to apply the fixes, waits for it, and reviews again. It stops when a review finds nothing or when it reaches the round limit, 10 by default. It asks for consent once per repository, refuses to run on the project's main checkout, and never pushes. Watch agent opens a live, read-only view of the fix agent while it works.","\u002Fdocs\u002Freview#review-and-fix-cycle",{"view":107,"tab":108,"anchors":137},[138],"review-cycle",{"id":140,"heading":141,"body":142,"state":104,"docsPath":143,"sim":144},"pr-conversation","The pull request conversation, under the diff","Once a session has a pull request on GitHub or Bitbucket, its review threads appear under the lines they are about. Comments, verdicts and outdated threads sit in a section above the diff. Reply, resolve, start a new thread, or publish one of your local notes to the pull request after a confirmation that names the repository. Send PR comments to agent hands the threads to the agent. Plexus fetches the conversation when you open the tab, focus the window or press Refresh. There is no background polling.","\u002Fdocs\u002Freview#pull-request-comments",{"view":107,"tab":108,"selected":145,"anchors":146,"note":149},"pricing-page-copy",[140,147,148],"comment-pr","send-comments","pricing-page-copy has an open PR with two PR comments.",{"id":151,"heading":152,"body":153,"state":104,"docsPath":154,"sim":155},"other-branches","Review a colleague's branch as well","Start a session on an existing branch, such as a colleague's branch or the head of a pull request you were asked to review, and set Compare with to the branch it will merge into. The Review tab then shows their work with the same diff, notes, AI review and pull request conversation. The session has its own worktree, so your main checkout and your other sessions stay where they are.","\u002Fdocs\u002Freview#reviewing-someone-elses-branch",{"view":107,"tab":108,"anchors":156},[157,110],"context",{"title":38,"description":78},"features\u002Freview","_qkQyPqrYApPNrlDa2oxofxv7yJcXq-W9Xc8QMnEqE8",[162,165,166,169,172,175],{"path":7,"title":163,"navTitle":6,"description":164},"Parallel Claude Code sessions with live status","Run many Claude Code agents at once, each in its own git worktree, and see which one needs you: working, needs you, delegating or idle, across every repo.",{"path":10,"title":38,"navTitle":9,"description":78},{"path":13,"title":167,"navTitle":12,"description":168},"A git worktree per agent, set up by your own hooks","Every Claude Code session gets its own git worktree, your .env files, a block of dev-server ports, and your Create, Run and Teardown hooks at three levels.",{"path":16,"title":170,"navTitle":15,"description":171},"Git client, PRs and GitHub issues for AI agents","A git client per Claude Code session and per repo, GitHub and Bitbucket pull requests with CI status, and GitHub Issues and Projects boards in one window.",{"path":19,"title":173,"navTitle":18,"description":174},"Stacked pull requests and ordered merging for Claude Code sessions","Stack Claude Code sessions on each other's branches. Plexus keeps every branch synced as the agents work, shows the stack as a map, and merges it in order, as a GitHub stack or through a merge queue.",{"path":22,"title":176,"navTitle":21,"description":177},"Teams: organizations, managed settings, prompts","Organizations, shared workspaces, managed settings, a shared prompt library and sync for teams on Claude Code. Seats are paid; the core app stays free.",1792695507458]