zoo-zero

Execute end-to-end tasks with iterative code-review and fix loops.

5|1|Updated Aug 22, 2025

Search Tags:#workflow #documentation #subagents #quality-assurance #code-review #bureau-mcp #zoo-zero

Ready to run

npx skills add https://github.com/andreyvit/zoo --skill zoo-zero

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: zoo-zero
Source: https://github.com/andreyvit/zoo/tree/main/.codex/skills/zoo-zero
Command: npx skills add https://github.com/andreyvit/zoo --skill zoo-zero

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Direct execution followed by structured review and fixes enables rapid end-to-end task completion without heavy upfront planning or task decomposition.

Core Features & Use Cases

End-to-end task ownership by a single agent, from research to delivery.
Integrated review and fix loops to surface issues and close tasks efficiently.
Suitable for coding, documentation, browser checks, testing, and research workflows.

Quick Start

Initiate Zoo Zero to perform the task end-to-end, then run the delegated code-reviews and iterations until the reviewer signs off.

Frequently Asked Questions about zoo-zero

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I handle end-to-end coding tasks without upfront planning artifacts?▼

End-to-end coding tasks without upfront planning artifacts can be handled by a single agent taking direct ownership, executing the work, and then running structured code-review iterations until the reviewer signs off.

What is the best way to automate review and fix loops for software development?▼

Automated review and fix loops are handled by delegating code-review mechanisms after direct task execution, allowing the agent to surface issues and close tasks efficiently without manual intervention.

Can I use a single agent workflow for documentation and research tasks?▼

Yes, a single agent workflow supports documentation, browser checks, and research tasks by performing end-to-end execution with direct control and iterative review cycles, ensuring comprehensive delivery.

Does this no-spec task execution approach require subagents or task decomposition?▼

No, the no-spec mode operates entirely without subagents or task decomposition, relying instead on a solo-agent workflow that executes tasks directly and uses review loops to ensure quality.

When should I avoid using a direct execution and review loop workflow?▼

You should avoid this direct execution workflow if your project requires heavy upfront planning, explicit task decomposition, or multi-agent orchestration rather than a solo-agent iterative approach.