clone-labs

C L O N E - L O O P

Keep Agent working, with predicted next prompts.
_{An agent plugin for self-driving coding loops powered by Clone MCP.}

Quick Start · How It Works · Commands · API Key · Versions

Clone Loop turns any agent task into a self-driving loop for Claude Code and Codex. When the agent tries to stop, Clone predicts your most likely next prompt and hands it back so the work continues — without you having to type the same nudge ten times.

/clone:loop "Build a REST API for todos. CRUD, validation, tests." --max-iterations 20

Then walk away.

Why people use it

• Stop typing "yes, keep going", "make test", "fix the failures", or "try again." Clone predicts the obvious next step and injects it for you. Threshold-gated so it only auto-continues when it's confident.
• Rich context, not a one-line snapshot. Each prediction sees your original task, every prior iteration's user turn, and a full timeline of what the agent did this iteration — text, tool calls, and tool results.
• AskUserQuestion popups answered automatically in Claude Code during an active loop.
• One session, one agent. No subprocesses, no daemons, no parallel agents to herd.

Quick start

Paste this into your agent after install:

/clone:loop "Run tests and fix any failures" --max-iterations 5

Install on macOS/Linux:

curl -fsSL https://raw.githubusercontent.com/cloneisyou/clone-loop/main/scripts/install.sh | bash

Install on Windows PowerShell:

irm https://raw.githubusercontent.com/cloneisyou/clone-loop/main/scripts/install.ps1 | iex

Or install manually from your shell:

claude plugin marketplace add cloneisyou/clone-loop@main
claude plugin install clone-labs@clone-loop --scope user

PowerShell:

claude.exe plugin marketplace add cloneisyou/clone-loop@main
claude.exe plugin install clone-labs@clone-loop --scope user

Open your agent and run:

/clone:api-key status
/clone:api-key login
/clone:interview "Clarify the feature before coding" --mode deep
/clone:loop "Run tests and fix any failures" --max-iterations 5

Cancel anytime with /clone:cancel-loop.

[!NOTE] Clone ships with a public demo API key so you can try it in seconds. For private memory and your own prediction quality, run /clone:api-key login.

To update later: claude plugin marketplace update clone-loop && claude plugin update clone-labs@clone-loop.

Commands

Command	What it does
/clone:interview "<topic>" [options]	Clarify a goal into an executable plan.
/clone:loop "<task>" [options]	Start a loop.
/clone:cancel-loop	Cancel the active loop.
/clone:api-key login|status|import-env|set|clear	Manage your Clone API key.
/clone:help	Show command help.

Options for /clone:interview

• --mode <quick|deep> — interview depth. Default deep.
• --max-questions <n> — maximum questions before restating. Default 12.
• --output <path> — project-local markdown spec path. Default .claude/clone-interview.local.md.
• --clone-threshold <n> — confidence threshold for Clone-predicted interview answers. Default 0.75.
• --no-auto-answer — disable Clone-predicted answers and always ask you.

Clone Interview is the goal-to-plan side of Clone. It inspects repo facts, asks the highest-impact unresolved question one at a time, asks Clone MCP to predict how you would answer, and only auto-records the answer when confidence clears the threshold. Low-confidence questions escalate to you.