Content
92%Weight 40%Scale 1-3Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is lean, actionable, and structured around a clear Plan-Execute-Verify loop with explicit validation. Its only meaningful weakness is a broken reference to best_practices.md, which is not present in the bundle.
Suggestions
Add the missing reference/best_practices.md file (and a references/ directory) so the two inline links resolve, or remove the references if the guidance fits inline.
Consider a one-line example of the write/search_replace edit step so the Execute phase is as concrete as the Plan and Verify phases.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is a lean list of concrete tool calls and brief rules with no padding or explanation of concepts Claude already knows; every token earns its place. | 3 / 3 |
Actionability | Provides specific executable MCP tool names (e.g., mcp_marimo_lint_notebook) and concrete rules (mo.sql, plt.gca(), "No plt.show()"), giving copy-paste-ready guidance. | 3 / 3 |
Workflow Clarity | The Plan-Execute-Verify loop is numbered and sequenced, with Verify marked "(CRITICAL)" and broken into lint, runtime, and output checks forming an explicit validation feedback loop. | 3 / 3 |
Progressive Disclosure | The body is a clear overview signaling a one-level-deep reference (reference/best_practices.md), but that file does not exist in the bundle (no references/ directory), so navigation is broken and cannot reach a 3. | 2 / 3 |
Total | 11 / 12 Passed |