com um clique
ralphi-loop
Executing a single iteration of the ralph autonomous coding loop. Use when ralph invokes the agent to implement the next user story from the PRD.
Menu
Executing a single iteration of the ralph autonomous coding loop. Use when ralph invokes the agent to implement the next user story from the PRD.
Scans a project and sets up ralphi loop configuration. Use when asked to initialize ralphi, set up a ralphi project, configure a project for AI-driven development, or run /ralphi-init.
Converts PRD markdown files to .ralphi/prd.json format for the Ralph autonomous agent loop. Use when asked to convert a PRD, create .ralphi/prd.json, or turn a PRD into Ralph format.
Generates structured Product Requirements Documents from feature descriptions. Use when asked to create a PRD, write requirements, plan a feature, or draft a spec.
| name | ralphi-loop |
| description | Executing a single iteration of the ralph autonomous coding loop. Use when ralph invokes the agent to implement the next user story from the PRD. |
You are an autonomous coding agent working on a software project.
Read .ralphi/config.yaml for project name, language, and framework.
Read the rules section in .ralphi/config.yaml and follow every rule listed.
Read the boundaries.never_touch section in .ralphi/config.yaml. Do NOT modify any file matching those patterns.
.ralphi/prd.json.ralphi/progress.txt (check Codebase Patterns section first)branchName before iteration starts).status: "open"status: "in_progress"feat: [Story ID] - [Story Title]status: "done".ralphi/progress.txtRun this protocol after implementation and before quality checks/commit.
.ralphi/prd.json (description + acceptance criteria)## Codebase Patterns and the latest iteration entry in .ralphi/progress.txt.ralphi/config.yaml rules and boundaries are still respectedReview Outcome to exactly one value: PASS or CHANGES_REQUIREDCHANGES_REQUIRED, fix issues and run this protocol againON_TRACK — Implementation is aligned with PRD intent/scopeRISK — Still aligned, but has a known dependency/open riskDRIFT — Current implementation diverges from PRD intent/scopeRISK or DRIFTDRIFT, define a corrective plan and do not mark the story complete until correctedRead the commands section in .ralphi/config.yaml and run each configured command (test, lint, build, typecheck, etc.). You can also run ralphi check which runs all commands in sequence.
Quality checks are also enforced by a pre-commit hook. If you try to commit and checks fail, fix the issues and try again. Do NOT use git commit --no-verify.
APPEND to .ralphi/progress.txt (never replace, always append):
## [Date/Time] - [Story ID]
Session: [Session URL/ID if available, else N/A]
Review Outcome: [PASS | CHANGES_REQUIRED]
Trajectory: [ON_TRACK | RISK | DRIFT]
Trajectory Notes: [Short rationale; required for RISK/DRIFT, else N/A]
Corrective Plan: [Required if DRIFT, else N/A]
- What was implemented
- Files changed
- **Learnings for future iterations:**
- Patterns discovered (e.g., "this codebase uses X for Y")
- Gotchas encountered (e.g., "don't forget to update Z when changing W")
- Useful context (e.g., "the evaluation panel is in component X")
---
Include a session reference if your engine provides one. Otherwise write N/A.
If you discover a reusable pattern, add it to the ## Codebase Patterns section at the TOP of .ralphi/progress.txt (create if it doesn't exist). Only add patterns that are general and reusable, not story-specific.
Before committing, check if edited files have learnings worth preserving in nearby AGENTS.md files:
.ralphi/progress.txt before starting