Content
38%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 buries a small amount of genuinely useful material (parameters table, output schema, one real command) under a large amount of duplicated, generic process boilerplate. It also contains factual inconsistencies with its own bundle — claiming 'No scripts included' while shipping scripts/main.py, and documenting CLI flags the script does not implement — plus a stale review date and non-portable example paths.
Suggestions
Cut the duplicated description text (repeated verbatim in 'When to Use' and 'Key Features'), the self-referential 'See ## Features/## Workflow above' lines, and the redundant 'Audit-Ready Commands' section that repeats the Quick Check; consolidate to one statement of purpose and one command block.
Delete or relocate the generic meta-boilerplate (Risk Assessment, Security Checklist, Lifecycle Status, Output Requirements, Response Template, Input Validation) — none of it is specific to FAQ generation, and the Risk Assessment's 'No scripts included / No external API calls' rows are factually wrong for this bundle.
Fix the accuracy of the executable guidance: remove the hardcoded '20260318/scientific-skills/...' cd path, align the documented parameters (--input/--audience/--output/--format) with what scripts/main.py actually accepts (or add argparse to the script), and link references/guidelines.md by name so the reference file is discoverable.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is heavily padded with generic meta-boilerplate — 'Output discipline: keep results reproducible, identify assumptions explicitly', the Risk Assessment / Security Checklist / Lifecycle Status sections, and a 7-item 'Response Template' — none of which is specific to FAQ generation. The full description is also pasted verbatim twice ('When to Use' bullet and 'Key Features' bullet), and the 'Quick Check' and 'Audit-Ready Commands' sections repeat the same py_compile command. It is not a 1 because there is no basic-concept tutorial padding (no 'what is an FAQ' explainer) and some sections (Parameters, Output Format) carry real information. | 2 / 5 |
Actionability | There is concrete guidance — a parameter table (--input/--audience/--output/--format), runnable commands ('python -m py_compile scripts/main.py', 'python scripts/main.py --help'), and a JSON output schema — but it is incomplete and partly unreliable: the example 'cd's into an environment-specific hardcoded path ('20260318/scientific-skills/Academic Writing/faq-generator'), the documented CLI flags do not exist in the packaged scripts/main.py (which takes no arguments), and the Risk Assessment row 'Code Execution: No scripts included' contradicts the actual bundle. This sits between 'minimal concrete guidance' (2) and 'mostly executable guidance' (4). | 3 / 5 |
Workflow Clarity | The Workflow section lists a 5-step sequence and does include checkpoints ('stop early if the task would require unsupported assumptions', 'If execution fails... switch to the fallback path'), plus a Quick Check compile step. However the steps themselves are abstract ('Use the packaged script path or the documented reasoning path') with no concrete commands embedded at each step, and the fallback path is never defined concretely. This matches 'steps listed but checkpoints missing or implicit / gaps' better than the clear-executable-sequence anchor at 4. | 3 / 5 |
Progressive Disclosure | The bundle has references/guidelines.md and scripts/main.py, but the body never names or links guidelines.md — it only gestures vaguely at 'Reference material available in references/' and 'references/ contains supporting rules, prompts, or checklists' — so the actual reference is buried. Meanwhile ~180 lines of boilerplate (security checklists, risk tables, lifecycle metadata, response templates) that belongs in separate files or nowhere are inlined in SKILL.md, and navigation is self-referential filler ('See ## Features above', 'See ## Workflow above' — including forward references to sections defined later). This matches 'references are buried / content that clearly belongs in separate files is inlined' rather than the better-organized anchor at 3. | 2 / 5 |
Total | 10 / 20 Passed |