Content
56%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 body is a well-organized, highly actionable SymPy reference with concrete executable examples and useful patterns, but it is substantially too verbose — duplicating tutorial content Claude already knows across redundant sections — and its progressive-disclosure design is unrealized: all five referenced files are absent from the bundle while their content is inlined anyway. Tightening the body to a true overview and shipping the reference files would fix both weaknesses.
Suggestions
Cut tutorial content Claude already knows and remove duplicated sections — the Overview (which repeats the description verbatim), 'Getting Started Examples' (re-covers Core Capabilities), and the 'Quick Reference' import list — keeping only SymPy-specific guidance like assumptions, exact-vs-float pitfalls, and solver selection.
Create the five promised reference files (references/core-capabilities.md, matrices-linear-algebra.md, physics-mechanics.md, advanced-topics.md, code-generation-printing.md) or remove the references to them, since the bundle currently contains no references/ directory and every link is broken.
Fix the non-executable snippets: import `Operator` and define `B` in the quantum mechanics example, remove `evalf` from the top-level import list (it is a method, not a module export), and add missing imports (`Integral`, `latex`, `pretty`, `sin`, `cos`) in the pattern examples.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~490-line body is a full SymPy tutorial covering textbook material Claude already knows (e.g. "diff(x**2, x) # 2*x", "factor(x**2 - 1) # (x - 1)*(x + 1)"), with several padded/redundant sections: the Overview repeats the frontmatter description verbatim, 'Getting Started Examples' re-covers the quadratic/derivative/eigenvalue examples already shown in Core Capabilities, and the 'Quick Reference' import list duplicates imports shown throughout. This matches anchor 2 (noticeably verbose, several unnecessary or padded sections); not 3 because the redundancy is systematic rather than incidental. | 2 / 5 |
Actionability | Guidance is overwhelmingly concrete: copy-paste-ready code with expected outputs in comments covering the common cases (solving, calculus, matrices, lambdify). Minor gaps keep it at anchor 4 rather than 5: the quantum mechanics snippet uses an undefined `B` and unimported `Operator`, `from sympy import evalf, N, nsimplify` includes `evalf` which is not a top-level SymPy export, and Pattern 3 uses `Integral`, `latex`, and `pretty` without importing them. | 4 / 5 |
Workflow Clarity | The 'Common Use Case Patterns' section provides a numbered 4-step symbolic-to-numeric pipeline and a solve-and-verify pattern with an explicit checkpoint ("assert result == 0"), and the troubleshooting section gives error-recovery guidance for common failures. This matches anchor 4 (clear sequence, most checkpoints present, minor validation gaps); not 5 because the pipeline pattern itself has no validation step. | 4 / 5 |
Progressive Disclosure | References are well-signaled one level deep ('For detailed calculus operations: See references/core-capabilities.md') with a 'Reference Files Structure' section giving per-file 'Load when' guidance — but the actual bundle contains no references/ directory, so all five referenced files are missing, and the body still inlines ~490 lines of the very content those files are described as covering. This matches anchor 3 (some structure but could be better organized); not 4 because the promised split between overview and reference files is not realized in the bundle. | 3 / 5 |
Total | 13 / 20 Passed |