> ## Documentation Index
> Fetch the complete documentation index at: https://recursive-mode.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# recursive-mode for agentic engineering

> A structured, file-backed workflow for coding agents. Give your agent audited phases, durable artifacts, and persistent memory — instead of losing everything in chat history.

recursive-mode is an installable skill package that transforms how AI agents handle complex engineering tasks. Instead of relying on conversational context that disappears between sessions, recursive-mode keeps requirements, plans, implementation evidence, and decisions in repository files — making agent work auditable, resumable, and consistent across contributors.

<CardGroup cols={2}>
  <Card title="Quick Start" icon="rocket" href="/quickstart">
    Install and bootstrap a repository in minutes
  </Card>

  <Card title="Installation" icon="download" href="/installation">
    All install options: skills CLI, Python, PowerShell, Bash
  </Card>

  <Card title="Workflow Overview" icon="diagram-project" href="/concepts/workflow-overview">
    Understand the phase-by-phase run model
  </Card>

  <Card title="Subskills" icon="puzzle-piece" href="/subskills/recursive-worktree">
    Specialized skills for TDD, debugging, review, and more
  </Card>
</CardGroup>

## An alternative to Missions

recursive-mode, previously known as rlm-workflow, is a free alternative that pre-dates Factory.ai's Missions feature by several months. It has a stronger recursion mechanism, is free and open source, and works in any IDE, CLI, agent and with any models.

## How it works

<Steps>
  <Step title="Install the skill package">
    Add recursive-mode to your agent environment with the skills CLI or a bootstrap script.
  </Step>

  <Step title="Bootstrap your repository">
    Run the installer to create the `/.recursive/` scaffold, control-plane docs, and memory layout.
  </Step>

  <Step title="Start a run">
    Give your agent a short command like `Implement the run` — requirements live in repo files, not in prompts.
  </Step>

  <Step title="Progress through audited phases">
    Each phase produces a locked artifact. The agent cannot skip phases or rewrite history — only add addenda.
  </Step>
</Steps>

## What's included

<CardGroup cols={2}>
  <Card title="Phases & Artifacts" icon="file-lines" href="/concepts/phases">
    Requirements, AS-IS analysis, planning, implementation, testing, QA, and closeout — all backed by locked files
  </Card>

  <Card title="Memory" icon="brain" href="/concepts/memory">
    Durable project knowledge persists across runs so agents start with better context every time
  </Card>

  <Card title="Scripts" icon="terminal" href="/scripts/overview">
    Cross-platform tools for status, linting, locking, review bundles, and hygiene checks
  </Card>

  <Card title="Guides" icon="book-open" href="/guides/starting-a-run">
    Step-by-step walkthroughs for common workflows and scenarios
  </Card>
</CardGroup>