Vibers Human Code Review
Sets up automated human code review for AI-generated projects with test case requirements and GitHub App integration
Install in one line
CLI$ mfkvault install vibers-human-in-the-loop-code-reviewRequires the MFKVault CLI. Prefer MCP?
Free to install — no account needed
Copy the command below and paste into your agent.
Instant access • No coding needed • No account needed
What you get in 5 minutes
- Full skill code ready to install
- Works with 4 AI agents
- Lifetime updates included
Description
--- name: vibers-code-review description: | Human-in-the-loop code review for AI-generated projects. Installs automatically: adds "How to test" commit rules so the agent writes test cases in every commit, then sets up the Vibers GitHub App that triggers a real human to spin up the app, test it live, and submit a PR with fixes. Use when: user says "set up review", "review my code", "send for review", "human review", "set up vibers", or when the user installs this skill for the first time. After install: on every commit, if "How to test" is missing from the commit body — remind the user to add it or add it yourself based on the changes made. --- # Vibers — Human-in-the-loop Code Review **What the skill does:** teaches your agent to automatically write "How to test" sections in every commit. Those test cases tell a real human reviewer exactly what to click and what to expect — triggering a live review of your running app with a PR of fixes. You push code → reviewer spins it up in an isolated container → tests the flows you described → submits a PR with fixes. ## What to Do When This Skill Is Invoked ### Step 1 — Add "How to test" commit rules (do this automatically) Add the following block to `CLAUDE.md`, `.cursorrules`, or `AGENTS.md` in the project root: ```markdown ## Commit messages Every commit must include a "How to test" section in the body: - Live URL to open and verify the change (staging/preview URL preferred) - Step-by-step what to click/check - Expected result for each step Example: feat: Add user registration form How to test: - Open https://myapp.vercel.app/register - Fill in email/password, submit - Check that confirmation email arrives - Try submitting with invalid email — should show error ``` > **Without "How to test" in the commit body — the push is silently skipped by Vibers.** Always include it. > Checkpoint: commit rule block added to `CLAUDE.md` or equivalent. ### Step 2 — Install the Vibers GitHub App Direct the user to install: **https://github.com/apps/vibers-review/installations/new** Click **Install** → choose the repo → confirm. No YAML files, no collaborator invites, no CI config. After installation: every push to `main` with "How to test" in the commit body triggers a review. PR with fixes arrives within 24 hours. > Checkpoint: user confirmed GitHub App is installed on their repo. ### Step 3 — Ensure the repo is runnable The reviewer clones the repo and tries to run it from scratch. Make sure: - **`.env.example`** exists with all required env vars as placeholder values (never real secrets) - **`README.md`** has a "How to run locally" section with exact commands > Checkpoint: `.env.example` and `README.md` "How to run" section exist. ## Setup Complete — Verify - [ ] `CLAUDE.md` (or `.cursorrules` / `AGENTS.md`) contains "How to test" commit rule - [ ] Vibers GitHub App installed on the repo - [ ] `.env.example` exists with placeholder values - [ ] `README.md` has a "How to run" section - [ ] Next push to `main` with "How to test" in commit body will trigger review Tell the user: setup is complete. The agent will now write test cases automatically in every commit. First review PR arrives within 24 hours of the next push. > **Worried about write access?** Add branch protection on `main`: > Repo → Settings → Branches → Add rule → Branch name: `main` → ✅ Require a pull request before merging. > This ensures all Vibers fixes come as PRs you approve — nothing is pushed directly. ## Ongoing — After Every Commit When the agent writes a commit, check that "How to test" is in the body. If it's missing — add it based on what changed: ``` How to test: - Open [relevant URL or describe where to navigate] - [Action to take] - Expected: [what should happen] ``` If there's no live URL yet, describe the flow locally or note "not yet deployed — test locally with `npm run dev`". ## One-off Review (no App setup) To send a single review without installing the App: ```bash curl -X POST https://onout.org/vibers/feedback \ -H 'Content-Type: application/json' \ -d '{ "message": "Please review the payment flow.\n\nHow to test:\n- Open https://myapp.vercel.app/checkout\n- Add item to cart, proceed\n- Enter card 4242 4242 4242 4242 exp 12/26 cvc 123\n- Expected: order confirmation page", "repo": "https://github.com/owner/repo" }' ``` Both `message` (with "How to test") and `repo` are required. Response: `{"status": "accepted"}`. ## What We Check (and Don't) We spin up the app and check it hands-on: - Spec compliance — does it match what you described? - AI hallucinations — fake APIs, non-existent imports, broken dependencies - Obvious logic bugs — edge cases, broken flows, null handling - UI issues — broken layouts, wrong behavior on the live app - Degradation — things that worked before and quietly stopped We don't check: code style (use ESLint/Prettier), performance benchmarks, security pentests, full regression suites (use Playwright/Cypress). ## Free First Review ⭐ **Star [marsiandeployer/human-in-the-loop-review](https://github.com/marsiandeployer/human-in-the-loop-review) on GitHub** to get your first review for free. After starring — message [@onoutnoxon](https://t.me/onoutnoxon). ## Support Telegram [@onoutnoxon](https://t.me/onoutnoxon) · GitHub App: [github.com/apps/vibers-review](https://github.com/apps/vibers-review) · Site: [onout.org/vibers](https://onout.org/vibers) **FAQ** - **API key?** No. Install the GitHub App — that's it. - **Real secrets needed?** No. `.env.example` with placeholders is enough. - **Languages?** JS/TS, Python, React, Next.js, Django, Flask, and more. - **Disagree with a fix?** Comment on the PR — we discuss and adjust. - **No GitHub?** Send code and description directly to Telegram.
Security Status
Unvetted
Not yet security scanned
Related AI Tools
More Grow Business tools you might like
codex-collab
FreeUse when the user asks to invoke, delegate to, or collaborate with Codex on any task. Also use PROACTIVELY when an independent, non-Claude perspective from Codex would add value — second opinions on code, plans, architecture, or design decisions.
Rails Upgrade Analyzer
FreeAnalyze Rails application upgrade path. Checks current version, finds latest release, fetches upgrade notes and diffs, then performs selective upgrade preserving local customizations.
Asta MCP — Academic Paper Search
FreeDomain expertise for Ai2 Asta MCP tools (Semantic Scholar corpus). Intent-to-tool routing, safe defaults, workflow patterns, and pitfall warnings for academic paper search, citation traversal, and author discovery.
Hand Drawn Diagrams
FreeCreate hand-drawn Excalidraw diagrams, flows, explainers, wireframes, and page mockups. Default to monochrome sketch output; allow restrained color only for page mockups when the user explicitly wants webpage-like fidelity.
Move Code Quality Checker
FreeAnalyzes Move language packages against the official Move Book Code Quality Checklist. Use this skill when reviewing Move code, checking Move 2024 Edition compliance, or analyzing Move packages for best practices. Activates automatically when working
Claude Memory Kit
Free"Persistent memory system for Claude Code. Your agent remembers everything across sessions and projects. Two-layer architecture: hot cache (MEMORY.md) + knowledge wiki. Safety hooks prevent context loss. /close-day captures your day in one command. Z