Skip to content

Conversation

@mattw0110
Copy link

UI you can use to generate and convert your PRDs using cursor-cli

@honestemployee2024
Copy link

@mattw0110 @snarktank I don't like that ralph.sh hardcodes the provider. Can we have a config setting (like .env) to allow people to pick AMP vs Codex vs Github Copilot vs Claude Code / new providers? We could default to something like AMP if a provider isn't specified in .env

…nce JSON conversion prompt with detailed user story requirements
…mmed output and nested structures, improving robustness in parsing agent responses
@mattw0110
Copy link
Author

Add windows support and updated PRD UI

@mattw0110
Copy link
Author

@mattw0110 @snarktank I don't like that ralph.sh hardcodes the provider. Can we have a config setting (like .env) to allow people to pick AMP vs Codex vs Github Copilot vs Claude Code / new providers? We could default to something like AMP if a provider isn't specified in .env

Great idea! i think i did see someone in another PR might of done this already.

…1, enhance branch handling, and update README with usage instructions
@mattw0110
Copy link
Author

feat: add support for multiple worker agents in ralph.sh and ralph.ps1, enhance branch handling, and update README with usage instructions.

--worker / -w flag to select AI backend (cursor or amp)

@snarktank
Copy link
Owner

Thanks for putting this together, but I'm trying to keep Ralph simple and focused. This PR adds a lot of scope (web UI, Cursor CLI, Windows support, multi-worker agents) that I'm not looking to take on.

For now I'm just supporting Amp and Claude Code (see #2). If you'd like to maintain a fork with these extra features, that's totally fine—just not something I want to merge into main.

Closing this, but appreciate the contribution!

@snarktank snarktank closed this Jan 16, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants