Content
71%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-structured, token-efficient, and genuinely actionable for an instruction-only skill, with strong templates and validation checklists. Its main defect is progressive disclosure: it points to two bundle files that do not exist, leaving the skill effectively monolithic with broken navigation.
Suggestions
Create the referenced 'references/narrative-frameworks.md' and 'templates/all-hands-template.md' files (or remove the references) so the 'See templates/all-hands-template.md for the full template' pointer in Step 5 resolves.
Remove the 'Keywords' section — it duplicates trigger terms already present in the frontmatter description and spends tokens without adding guidance.
Add a short pass/fail instruction to the Narrative Consistency Checklist (e.g., what to do when a box fails) to close the validate→fix→retry loop hinted at in Step 3.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense with domain-specific judgment material (translation matrix, good/bad core narrative pair, failure modes) that Claude cannot be assumed to know, and avoids explaining generic concepts. Minor trimmable excess remains — the 'Keywords' section duplicates the frontmatter description's trigger terms, and a few aphorisms ('The good version is usable. The bad version says nothing.') could be compressed. Efficient with minor over-explanation matches anchor 4. | 4 / 5 |
Actionability | Concrete, executable artifacts dominate: a fill-in core narrative template, a good/bad worked example, an audience translation matrix, an hour-by-hour crisis sequence, and a copy-paste crisis communication template. Not 5 because guidance like 'Give people a reason to be proud of their choice to work here' and the cadence table stay at the principle level without worked outputs for each audience. Mostly executable guidance with minor gaps matches anchor 4. | 4 / 5 |
Workflow Clarity | The six steps are clearly sequenced (build core narrative → translate per audience → detect contradictions → cadence → all-hands → crisis) and the contradiction check plus the 'Narrative Consistency Checklist' serve as explicit validation checkpoints, with correction guidance ('Fix the less accurate version, then communicate the correction'). Not 5 because there is no full validate→fix→retry feedback loop and the checklist lacks explicit pass/fail handling; no destructive/batch cap applies since this is communication guidance. | 4 / 5 |
Progressive Disclosure | The body references 'references/narrative-frameworks.md' and 'templates/all-hands-template.md' in a 'Detailed References' section, but neither file exists in the bundle — the references are dangling, so navigation fails. Additionally, Step 5 defers to the missing all-hands template while inlining cadence/failure-mode content that belongs in that separate file. Some structure present with problematic references matches anchor 3; not 2 because the body itself is well-sectioned and the references are clearly signaled rather than buried. | 3 / 5 |
Total | 15 / 20 Passed |