Content
87%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-structured, concise skill that provides clear actionable instructions for setting up a Ralph loop. Its main weakness is the lack of validation steps—there's no guidance on verifying the scratchpad was written correctly, handling stop hook failures, or recovering from malformed state. The guardrails section is a nice touch for safety but could benefit from more explicit error handling.
Suggestions
Add a validation step after writing the scratchpad file (e.g., read it back and verify the YAML frontmatter parses correctly before confirming the loop is active)
Add error recovery guidance for when the stop hook doesn't fire or the scratchpad becomes corrupted mid-loop
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The content is lean and efficient. It doesn't explain what iterative loops are or how agents work—it assumes Claude's competence. Every section serves a clear purpose with no padding or unnecessary context. | 3 / 3 |
Actionability | Provides a concrete file format with exact markdown structure, a specific file path (.cursor/ralph/scratchpad.md), a worked example, and clear parameters with defaults. The instructions are copy-paste ready and unambiguous. | 3 / 3 |
Workflow Clarity | The steps are clearly sequenced (gather params → create file → confirm → begin work), but there's no validation checkpoint after writing the state file, and no error recovery guidance if the stop hook fails or the scratchpad format is malformed. For an iterative autonomous loop, missing feedback/validation steps is a notable gap. | 2 / 3 |
Progressive Disclosure | For a skill under 50 lines with no need for external references, the content is well-organized into clear sections (Trigger, Workflow, Guardrails, Output). The structure is appropriate for the skill's complexity. | 3 / 3 |
Total | 11 / 12 Passed |