Content
48%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The skill contains a solid executable core — precise data-format tables, an exact Rscript invocation, and a well-defined output template — but wraps it in substantial generic template boilerplate and contradictory guidance about the Python and R entry points, including a fabricated validation section. Tightening the document to the workflow, data requirements, invocation, and failure modes would substantially improve it.
Suggestions
Delete or drastically trim the generic boilerplate sections ("When to Use", "Key Features", "Implementation Details", "Required Inputs", "Output Contract", "Validation and Safety Rules", "Failure Handling", "Quick Validation") and keep only funnel-plot-specific guidance.
Resolve the entry-point contradiction: commit to either the R or the Python implementation in the Workflow, and remove the fabricated "Quick Validation" expected output that `funnel_plot.py --help` does not actually produce.
Move invocation details and package installation into (or adjacent to) the Workflow section so a reader follows one coherent path from input validation to output, and fix the "See `## Workflow` above" reference and the dated `cd "20260316/..."` example path.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is noticeably verbose: roughly half of it ("When to Use", "Key Features", "Implementation Details", "Required Inputs", "Output Contract", "Validation and Safety Rules", "Failure Handling", "Quick Validation") is generic template boilerplate unrelated to funnel plots, and "Implementation Details" opens with a broken self-reference ("See `## Workflow` above" when Workflow appears later). The genuinely useful content (data format tables, Rscript invocation, output template, package dependencies) is buried among padded sections. | 2 / 5 |
Actionability | The core path is concrete and executable — the exact Rscript command with per-parameter descriptions, column-name tables per data type, and an R package install command — but it is undermined by incoherent guidance: "Example Usage" and "Quick Validation" direct to `python scripts/funnel_plot.py --help` with a fabricated expected output (the script takes positional args and prints a usage line, not the documented result-file summary), while the actual workflow requires `Rscript scripts/funnel_plot.R`; the "Output Contract" also invents a `meta_funnel_plot_result.md` deliverable that contradicts the documented PNG/CSV outputs. These are more than minor gaps, so it cannot reach the 'mostly executable' anchor of 4. | 3 / 5 |
Workflow Clarity | The Workflow section gives a clear 3-step sequence (validate input → execute script → report output) with an explicit validation checkpoint (required columns per data type, minimum 3 studies) and a fully specified output template. It falls short of 5 because there is no in-workflow error-recovery loop (failure handling is only covered generically in a separate section) and the surrounding sections introduce Python-vs-R entry-point confusion. | 4 / 5 |
Progressive Disclosure | The body is sectioned with headers and the referenced bundle files (`scripts/funnel_plot.py`, `scripts/funnel_plot.R`) are real, but everything lives inline in a ~200-line SKILL.md with no reference files. The generic boilerplate sections inflate the main file without adding navigational value, and the confusing placement of content (workflow-relevant details in "Implementation Details" and "Example Usage" rather than the Workflow section) means organization could be better. | 3 / 5 |
Total | 12 / 20 Passed |