Skip to content
/ dalph Public
forked from snarktank/ralph

Dalph is fork of Ralph, but for RooCode. Autonomous AI agent loop that runs repeatedly until all PRD items are complete.

Notifications You must be signed in to change notification settings

theyv/dalph

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

12 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Dalph

image

Dalph = Dave + Ralph + Roo Code 🤖

Autonomous AI agent loop for Roo Code that runs repeatedly until all PRD items are complete. Fork of Geoffrey Huntley's Ralph, adapted for Roo Code's custom modes and tool system.

⚠️ This is NOT for token savers. This is for people who want stuff done autonomously (more or less). Expect high token usage - each story gets fresh context.

🖥️ Tested primarily on Windows with Roo Code. Should work on Mac/Linux but YMMV.


What's Different from Ralph?

Ralph (Amp):

  • Bash script loop
  • jq for JSON parsing
  • Single prompt.md
  • Re-analyzes codebase each iteration
  • Deletes test utilities
  • Files loose in /tasks/

Dalph (Roo Code):

  • Roo Code custom modes
  • Native Roo Code tools (Read/Edit Files)
  • Separate Orchestrator + Implementer modes
  • Merged PRD Creator mode + PRD JSON conversion
  • Analyzes once, caches in context.md
  • More "test focused"
  • Preserves tests in test_tools/ folder
  • Per-task folders: /tasks/[task-name]/

Flow

  1. dalph-prd-creator - Ask clarifying questions, generate PRD markdown, convert to prd.json

  2. dalph-orchestrator - Detect stack ONCE and create context.md, pick next story from prd.json, create subtask for implementer, process result and update prd.json, repeat until all stories pass or 3+ blocked

  3. dalph-implementer (called as subtask per story) - Read context.md (short file, not whole codebase), implement ONE story, verify each acceptance criterion, save test tools to test_tools/, commit and report back


Setup

1. Install Custom Modes in Roo Code

Go to Roo Code → Settings → Modes → Create New Mode and add three modes:

  • 🏭 DALPH: PRD Creator (slug: dalph-prd-creator) - see dalph-prd-creator/dalph-prd-creator-info.md
  • 🏭 DALPH: Orchestrator (slug: dalph-orchestrator) - see dalph-orchestrator/dalph-orchestrator-info.md
  • 🏭 DALPH: Implementer (slug: dalph-implementer) - see dalph-implementer/dalph-implementer-info.md

Copy the Role Definition, Short Description, When to Use, and Custom Instructions from each file.

2. Create 'tasks' Folder in your project directory

Usage

Step 1: Create PRD

Switch to dalph-prd-creator mode:

Create a PRD for [describe your feature]

Answer the clarifying questions (format: "1A, 2B, 3C"), review generated PRD, confirm JSON conversion when asked.

Output structure:

tasks/[task-name]/
  prd-[task-name].md
  prd.json
  progress.txt

Step 2: Run Orchestrator

  • Switch to dalph-orchestrator mode.

  • Set "BRRR" / "YOLO" mode ON to enable autonomous execution.

  • Run the orchestrator with your task:

run @/tasks/[task-name]/prd.json - (task: @/tasks/[task-name]/prd-[task-name].md )

Orchestrator will detect project stack, create context.md, then loop through stories - creating implementer subtasks, processing results, updating prd.json until done or 3+ stories blocked.

(Step 2.5 - go grab a coffee while it works. Or two. Maybe three.)

Step 3: Done

When all stories pass: <promise>COMPLETE</promise>

If blocked: <promise>BLOCKED - 3+ stories failed</promise>


Task Folder Structure

tasks/[task-name]/
  prd-[task-name].md   - PRD document (human-readable)
  prd.json             - Stories + progress (machine-readable)
  progress.txt         - Implementation log + learnings
  context.md           - Short context for implementers (~500 tokens)
  test_tools/          - Preserved test utilities

Key Concepts

Small Stories

Each story must be completable in ONE iteration. Split big stories. Good: "Add status column to database". Bad: "Build entire notification system".

Acceptance Criteria = Tests

Each criterion must be verifiable. Implementer will actually test each one. Good: "Button shows confirmation dialog before deleting". Bad: "Works correctly".

Context Efficiency

Orchestrator analyzes stack once, saves to context.md. Implementers read only context.md (~500 tokens), not whole codebase. New patterns flow back through orchestrator to context.md.

Test Tools Preserved

Implementers save verification scripts to test_tools/. These are kept for debugging, not deleted.


Checking Progress

cat tasks/[task-name]/prd.json      # story status
cat tasks/[task-name]/progress.txt  # implementation log
cat tasks/[task-name]/context.md    # discovered patterns
git log --oneline -10               # git history

Tips

  • Use it with Roo Code Cloud to grab more coffees while it running.

Credits


License

MIT

About

Dalph is fork of Ralph, but for RooCode. Autonomous AI agent loop that runs repeatedly until all PRD items are complete.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published