👁️ reslop
Review AI generated code and plan changes
Turn generated changes into owned changes
Open a change, leave findings and a backlog, and prepare a repair plan for the agent.
Review → Plan → Repair → Verify- Live dashboard: files, diffs, tasks, branches, commits, run, agents, and npm, updated in real time
- Leave feedback, a project backlog, and in-place code proposals as a repair plan
- Review uncommitted git diffs, a given commit, a GitHub PR, or a GitLab MR
- File-scope review mode: not just diffs, with editor mode
- Stage, unstage, or revert each contiguous block of diff lines
- Auto-reload local diffs when files change, keeping the current screen
- Import GitHub PR and GitLab MR review comments, and GitHub/GitLab issues, into tasks and the local plan
- Commits list: brief and full, view a commit's diff, commit, update, amend, apply, edit, fixup, and delete
- Branches: checkout, create, rebase, delete, pull, and push
- npm scripts and bins: run, edit, and reorder; double-click to run; clean logs older than 5 days
- Reduced test and lint output for agents: failures and a short summary, not passing noise
- Review, renew, audit and update npm dependencies in
package.jsonand the lockfile - Auto-update reslop patch and minor releases; confirm a new major
- Work with git, npm, and fetch in the background; run tests in the background while you review
- Execute the plan in local agents over ACP: pick agent, model, and effort; watch progress; sessions, history, and continue or new runs
- Agent defaults in
.reslop(grouped settings sections) - Optional line numbers in diff and file views (
Ctrl+L)
Install
npm i -g reslopWorks on Linux, macOS, and Windows. On Windows use Windows Terminal.
Requires Node.js >=18.15.0.
Usage
resloplive dashboard of this repository; press a block key to open its screenreslop foldertiles of the repositories in a folder, each with its git worktrees- git worktrees of a repository are tiles above its branches when there are several:
←→select a tile,↑↓go through the branches,enteror a click shows that worktree,insertanddeletecreate and remove one reslop path/fileshow only that path or file (starts on the file list)reslop 7ac260cshow that commit (read-only)reslop https://github.com/metarhia/metacom/pull/555GitHub/GitLab PR/MRreslop https://github.com/metarhia/metacom/issues/550import GitHub/GitLab issuereslop https://github.com/metarhia/metacom/issuesimport all issuesreslop -rread-only modereslop t --run one program and print a reduced report[args]
reslop t --. Keep && between those wraps, so each program is captured on its own and a failing step does not start the next one:
"lint": "reslop t -- eslint . && reslop t -- prettier -c \"*/.js\"",
"test": "reslop t -- npm run -s lint && reslop t -- node --test"The report is Markdown: failures, diagnostics, and a short summary. Passing results are omitted and similar problems are grouped. The reduced report is saved as .log/ and the raw output as .log/. The report names the raw file for an agent to read when a detail is missing. Set RESLOP_OUTPUT=raw to pass the command through unchanged. The review screen runs scripts in raw mode and applies its own filter once.
npm t runs lint and the tests that do not start a process. Run a test file that starts git, node, or npm with node --test . New reviews include both commands in the agent instructions.
reslop t records each run as .log/ beside those files so that the dashboard can show commands started by an agent in another terminal.
Reviews go in .plan/YYYY-MM-DD-NN.md with frontmatter status:
editing: writing the review in reslop; AI works through the checkboxespartial: AI started; some items remaindone: all items marked[x]; adding an issue in reslop sets it back toediting
ready and pending in older reviews are read as editing.
Agents
reslop runs a plan in an agent over the Agent Client Protocol. An agent is listed as available when the command that serves ACP is on PATH:
| Agent | Command | Install |
| ----------- | -------------------- | ------------------------------------------------ |
| Claude Code | claude-agent-acp | npm i -g @agentclientprotocol/claude-agent-acp |
| Codex | codex-acp | npm i -g @agentclientprotocol/codex-acp |
| Cursor | cursor-agent acp | comes with the Cursor CLI |
| OpenCode | opencode acp | npm i -g opencode-ai |
| Antigravity | agy_acp_server.par | downloaded from the official ACP registry |
A missing agent shows what it needs, in the warning color; i on that row proposes installing it. npm adapters are installed globally, while Google Antigravity's official ACP runtime is downloaded from the ACP registry into reslop's user data directory. reslop also discovers that runtime on PATH and in Zed's registry directory. After confirmation, reslop lists the agent as soon as the install finishes and does not restart. docs/acp-adapters.md describes the integrations.
Models are the ones the agent lists; efforts, context sizes, and fast mode are the ones it offers for the picked model. When a run ends, the session is set back to what it was, so the defaults of the agent's own CLI stay as they were. A run has nobody to ask, so reslop answers permission requests itself: edits inside the repository and npm, npx, node, reslop, and git commands, piped through read-only filters such as grep or tail, are allowed; everything else is refused.
Future
- Search, blame, and file history in the review
- Security and code-quality audit; propose a repair plan
- Send anonymized code blocks for expert review
- Send questions and the repair plan to experts for approval
- Ask experts
- Apply refactoring skills
- Squash, cherry-pick, and interactive rebase
- Merge, rename, and conflict resolution
License
Copyright (c) 2026 Timur Shemsedinov and other contributors (see github). This is MIT licensed software.