Content
52%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 well-organized with concrete output formats and a clear first-activation workflow, but it is undermined by a bundle that does not exist: the core Step 2 depends on reading `references/cheat-codes.md`, which is absent, along with all three referenced scripts. A duplicated README-style tail (Name, Description, Usage, Features) restates frontmatter and earlier sections, inflating token cost without adding guidance.
Suggestions
Ship the referenced bundle files (`references/cheat-codes.md`, `references/coaching-rules.md`, and the three `scripts/*.py`) or remove the references and inline the minimum needed content — the first-activation flow's Step 2 cannot execute without `references/cheat-codes.md`.
Delete the duplicated README tail ("## Name", "## Description", "## Features", "## Usage", "## Examples", "## Scripts" sections) — it restates frontmatter and repeats the activation phrases and examples already covered, cutting the body by roughly a third.
Fix the frontmatter hygiene: duplicate keys with conflicting values (`Name`/`name` and `Version: 1.0.0` vs `version: 2.9.0`) and unexplained metadata (`Tier: POWERFUL`, `Category: meta`) invite parsing ambiguity and drift.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The rules core (Rules 1-7, tone guidance) is tight, but the tail duplicates material: a "## Name" / "## Description" section restating frontmatter, "## Usage" re-listing the same activation phrases already given under "When to invoke this skill" ("Coach me on Claude", "Make me a Claude power user" appear twice), and "## Features"/"## Examples" re-describing what earlier sections already prescribe. Mostly efficient but clearly could be tightened by roughly a third; it is not anchor 2 because the padding is duplication rather than concept explanation, and the core sections earn their tokens. | 3 / 5 |
Actionability | Concrete, copy-paste-ready formats exist (the ⚡ tip template, the structured "rate that prompt" block, the first-activation 3-step sequence), but the pivotal Step 2 instructs "Read `references/cheat-codes.md`" and three `scripts/*.py` files are listed — none of which exist in the bundle — so the personalized glossary, the skill's core deliverable, cannot be produced as instructed. Not anchor 4 because key executable material is missing; not anchor 2 because substantial specific guidance and exact output formats are present. | 3 / 5 |
Workflow Clarity | The first-activation flow is a clearly sequenced 3-step process with an explicit conditional checkpoint ("If the user already mentioned their use case in the activating message, skip this question and proceed"), and ongoing behavior is enumerated as Rules 1-7 with explicit do/do-not trigger lists. No destructive or batch operations, so no validation loop is required. Not anchor 5 because there are no error-recovery or feedback checkpoints (e.g., what to do if the glossary cannot be built or the user pushes back mid-flow — the opt-out appears only in the Features tail). | 4 / 5 |
Progressive Disclosure | The body cites `references/cheat-codes.md` ("Read on first activation"), `references/coaching-rules.md`, and three scripts — but the bundle contains no references/, scripts/, or assets/ directories, so every referenced path is a dead end and the claimed file split is undelivered. The design intent is good (one-level-deep, clearly signaled references), but scoring against the actual bundle structure the navigation fails outright, and the README-style tail inlines frontmatter content (Name/Description) that does not belong in the body. Not anchor 1 because references are clearly signaled, one level deep, and not nested. | 2 / 5 |
Total | 12 / 20 Passed |