Content
87%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The content is lean, actionable, and well-structured for a simple skill, with concrete paths and a ready-to-use frontmatter example. The main gap is the absence of an explicit validation/verification checkpoint in the setup workflow for an autonomous loop.
Suggestions
Add a verification step after writing the state file (e.g., re-read .cursor/ralph/scratchpad.md to confirm the frontmatter parsed and the prompt is present) before declaring the loop active.
Include a brief note on how to detect/handle a malformed or missing state file when the stop hook fires, giving the loop a clear error-recovery path.
Clarify the termination condition more explicitly — e.g., state that the loop ends only when a truthful <promise> is emitted or max_iterations is reached — as a checklist item alongside the existing guardrails.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean (under 50 lines), assumes Claude's competence, and avoids explaining concepts Claude already knows; every section earns its place without padding. | 3 / 3 |
Actionability | Gives the exact state-file path (.cursor/ralph/scratchpad.md), the precise YAML frontmatter format, named parameters with defaults, and a copy-paste-ready example block. | 3 / 3 |
Workflow Clarity | Steps are clearly numbered (1-4) with the stop-hook feedback mechanism described, but there is no explicit validation checkpoint confirming the state file was written correctly before the loop begins, leaving checkpoints implicit for an autonomous iteration process. | 2 / 3 |
Progressive Disclosure | A simple, single-purpose skill under 50 lines with well-organized sections (Trigger, Workflow, Guardrails, Output) and no need for external references, satisfying the simple-skill allowance. | 3 / 3 |
Total | 11 / 12 Passed |