ralph

ralph is a cursor rule for coding agents from Jkudjo/oh-my-cursor. It costs 11 tokens per session (437 once invoked), scanned A, original, MIT.

A persistent execution loop for carrying out an approved coding plan one step at a time. It verifies each step and investigates the underlying cause when something fails.

In plain words
What is it for?
Use it to implement a feature or fix from a plan, run available tests and smoke checks after each step, and produce a final completion summary.
Why use it?
It prevents planned work from being skipped or marked complete while tests or other checks are still failing.

Cursor rule

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add rules/jkudjo/oh-my-cursor/ralph
Clone the repo
git clone --depth 1 https://github.com/Jkudjo/oh-my-cursor

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for ralph

README.md
[![agentmods](https://agentmods.dev/badge/rules/jkudjo/oh-my-cursor/ralph.svg)](https://agentmods.dev/rules/jkudjo/oh-my-cursor/ralph)
Your own site
<a href="https://agentmods.dev/rules/jkudjo/oh-my-cursor/ralph"><img src="https://agentmods.dev/badge/rules/jkudjo/oh-my-cursor/ralph.svg" alt="Measured on agentmods" height="20"></a>
Per session 11 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 437 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00011 $0.00437
Opus 5 $0.00005 $0.00218
Sonnet 5 $0.00002 $0.00087
Haiku 4.5 $0.00001 $0.00044

Measured 3d ago against content hash 2519aa1f3939, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

ralph scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

rules/ralph.mdc · 41 lines

What it actually says

@ralph

Trigger: After @ralplan has produced an approved plan, or for direct execution of a clear task.

Protocol

  1. Call start_mode with mode=ralph.
  2. Load the latest approved plan: call list_plans, then load_plan with the most recent approved plan ID.
  3. Execute each step in order:
    • Call update_mode_state with the current step as phase and increment iteration.
    • Implement the step completely before moving to the next.
    • After each step: verify it works (run tests if available, read output, check for errors).
    • If a step fails: diagnose the root cause, fix it, verify again. Do NOT skip.
  4. After all steps are complete:
    • Run a final verification pass (tests, linting, smoke check).
    • Call update_mode_state with status=completed.
    • Write a completion summary:
      ## Completed
      Steps done: X/X
      Files changed: ...
      Verification: passed / failed (details)
      

Rules

  • One step at a time. Complete and verify before moving on.
  • Never skip a failing step — fix the root cause, not the symptom.
  • Root cause before fix. Never apply a fix without stating what the root cause is. "It was failing" is not a root cause.
  • Never mutate tests to make them pass. If a test is wrong, flag it explicitly and get confirmation before changing it. Silencing assertions is not fixing.
  • Minimum viable change. Fix exactly what is broken. Do not refactor surrounding code while fixing a bug.
  • If blocked: call update_mode_state with phase=blocked and explain the blocker clearly.
  • Persist progress: call update_mode_state after each step so state is recoverable.
  • Blast radius. Before each step, state what could break if the change is wrong.
Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 3d ago First seen · 41 lines · 11 tokens per session scan A 2519aa1f3939

Subscribe to this mod's changes

ralph is a cursor rule published in the GitHub repository Jkudjo/oh-my-cursor (1 stars, last pushed 5mo ago), licensed MIT. It adds 11 tokens to every session and 437 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other cursor rules, from other repositories