AgentIndex icon
AgentIndex
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Submit Tool
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Home/
Multi-Agent/
zeroshot
zeroshot logo

zeroshot

Active·★ 2.0k·MIT·Updated 2026-10-27
★ Essential★ Trending

Zeroshot is an open-source CLI that orchestrates AI coding agents through a plan, implement, validate and iterate loop. A planner turns a task into acceptance criteria, an implementer makes the change in an isolated workspace (local, a git worktree or Docker), and independent validators check the result without seeing the implementer's context or history, repeating until the change is verified or rejected with actionable findings. State is stored in SQLite so a crashed run can resume, and model ceilings are meant to prevent runaway API spend. Tasks can come from a GitHub issue number, a markdown file or an inline prompt, with GitHub, GitLab, Jira and Azure DevOps as issue backends.

Install with `npm install -g @covibes/zeroshot` on Linux or macOS (Windows WSL is not yet supported). It needs Node 18 or newer and at least one provider CLI, such as Claude Code, Codex, Gemini or OpenCode, authenticated on your machine, and `gh auth login` if you pass issue numbers. `zeroshot run 123` or `zeroshot run "Add dark mode"` starts a run, and `zeroshot providers set-default codex` picks the default provider. The README says it works best when a task has clear acceptance criteria and gives a table of good and bad fits, such as fixing a login bug versus "make the app faster". It is MIT licensed.

This write-up is based on the project's README and public repository data. We have not hand-tested the tool. Data as of 2026-10-25.

License
MIT
Latest release
v10.18.0 · 2026-10-24
Last commit
2026-10-25
Contributors
22
Forks
195
Open issues and PRs
12
First published
2025-12-25

Source: GitHub, npm and PyPI public APIs.

#AI Agent#Code Automation#CLI#Multi-Agent System#DevOps#Coding
$ Install
$ npm install -g @covibes/zeroshot
↗ Visit site★ GitHub
01

Features

01Blind validation with isolated agents
02Repeatable multi-agent workflows
03Accept/reject loop with actionable findings
04Crash-safe state persisted for recovery
05Multiple code isolation modes (worktree, Docker)
02

Why choose it

+Independent validators that never see the implementer's context, which makes the check more meaningful than self-review
+Isolation modes: local, git worktree or Docker
+Crash-safe state in SQLite and model ceilings to limit API spend
+Works with Claude Code, Codex, Gemini or OpenCode, and with GitHub, GitLab, Jira and Azure DevOps issues
03

Trade-offs

!Multiple agents looping until approval can use many tokens; the README's demo is a 90-minute, 5-iteration run
!Works best only when acceptance criteria are clear; exploratory tasks are listed as poor fits
!Needs a separately installed and authenticated provider CLI
!Linux and macOS only; Windows WSL is not supported yet
04

Maintenance check

ActiveChecked: 2026-10-26
Last commit
2026-10-26
Commits, last 90 days
100+
Releases, last 90 days
30
Latest release
v10.19.1 (2026-10-26)
Open issues
10
Recent issues with at least one comment
3/8
Commits per week, last 12 weeks
Commits per week, last 12 weeks: 40, 8, 9, 7, 26, 31, 28, 9, 33, 28, 33, 5

Oldest week on the left; the latest four weeks are orange. Source: GitHub commit activity.

Status rule: archived on GitHub = Archived; last commit within 90 days = Active; 91 to 365 days = Slowing; older = Stale. All figures come from the GitHub API on the date shown.

Most discussed open issues

  • #1301 [BUG] Local delivery ignores global Git core.autocrlf configuration2 comments · 2026-10-24
  • #1203 [BUG] Copilot grammar-compilation rate limits fail the run instead of backing off1 comments · 2026-10-07
  • #1299 [BUG] Copilot and capsule test fixtures hardcode /usr/bin/python3 and cannot exec where Python lives elsewhere0 comments · 2026-10-24
  • #1210 [FEATURE] Gitea git forge setup0 comments · 2026-10-07
  • #1206 [BUG] A reviewer node can wipe the worktree with destructive Git commands0 comments · 2026-10-07

Issue titles are quoted from GitHub. Open the link to check whether a problem still applies to the version you use.

05

Compatibility

Claude
Supported AI
Verified via docs
Codex
Supported AI
Verified via docs
Gemini
Supported AI
Verified via docs
OpenCode
Supported AI
Verified via docs
GitHub
Issue Backend
Verified via docs
GitLab
Issue Backend
Verified via docs
06

Quick start

1
$ npm install -g @covibes/zeroshot
07

Use cases

↳Autonomous AI code refactoring
↳AI-powered pull request automation
↳Automated bug fixing with validation
08

Alternatives

⚡Top VS:
VSMetaGPT★71.0k→VSintentkit★6.5k→VSii-agent★3.4k→VSCodeMachine-CLI★2.5k→
MetaGPT logo
MetaGPT★ 71.0k
🌟 The Multi-Agent Framework: First AI Software Company, Towards Natural Language Programming
VSCompare→
intentkit logo
intentkit★ 6.5k
An open and fair framework for everyone to build AI agents equipped with powerful skills. Launch your agent, improve the world, your wallet, or both!
VSCompare→
ii-agent logo
ii-agent★ 3.4k
II-Agent: a new open-source framework to build and deploy intelligent agents
VSCompare→
CodeMachine-CLI logo
CodeMachine-CLI★ 2.5k
CLI-native orchestration engine for autonomous workflows and production-ready code. ⚡️
VSCompare→
agentUniverse logo
agentUniverse★ 2.4k
agentUniverse is a LLM multi-agent framework that allows developers to easily build multi-agent applications.
VSCompare→
thClaws logo
thClaws★ 1.2k
Open-source AI agent harness in native Rust — GUI, CLI, headless, and webapp from one binary. Multi-provider, MCP, skills, plugins, agent teams.
VSCompare→
metaharness logo
metaharness★ 716
🛠️ The meta-harness for AI agents — scaffold your own focused, branded agent harness with its own npx CLI, MCP server, memory, learning loop, and witness-signed releases. Works with Claude Code, Codex, pi.dev, Hermes, OpenClaw, and RVM (hardware-isolated sandbox).
VSCompare→
BettaFish logo
BettaFish★ 42.5k
Multi-agent opinion analysis assistant. Breaks information bubbles and predicts trends.
VSCompare→
See all alternatives →

Comments

Log in to leave a comment

No comments yet. Be the first!

On this page
01Features02Why choose it03Trade-offs04Maintenance check05Compatibility06Quick start07Use cases08Alternatives
covibes /
zeroshot
↗

CLI that runs planner, implementer and blind validators until code is verified

2.0k195MIT
Last commit1d ago
StatusActive
CategoryMulti-Agent
Contributors22
Latest releasev10.18.0
Links
Documentation↗Discussion↗Issues↗Releases↗
Cloud Infrastructure
Ad

Deploy on DigitalOcean

Host your MCP servers, Docker containers, and AI agents with $200 free credit over 60 days.

Claim $200 Free Credit
© 2026 AgentIndex.app|Built by a 10-year iOS Developer.
Best by use caseAboutContactAdvertisePrivacy PolicyTerms of UseQYSGitHubBuy me a coffee ☕

Browse by Category

MCP ServersCode AssistantWorkflow AutomationRAG / Knowledge BaseMulti-AgentBrowser AutomationLLM InfraDev ToolingObservability

Not affiliated with Anthropic, OpenAI or Microsoft.