Content
78%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 lean, highly actionable playbook — concrete tables, thresholds, workshop questions, and stage-specific rituals organized in a coherent build→measure→maintain lifecycle. Its weaknesses are the broken reference chain (both cited bundle files are absent from the skill directory) and minor token redundancy from a duplicated Keywords section and repeated reference pointers.
Suggestions
Ship the referenced bundle files or inline their essentials: `references/culture-playbook.md` and `templates/culture-code-template.md` are cited three times but do not exist in the skill directory — either add them (with the survey questions and culture code template) or inline the critical content so the progressive disclosure chain is not broken.
Remove the `## Keywords` section and the `## Detailed References` section: the frontmatter description already carries the trigger terms and each framework already cites its reference inline, so these ~5 lines are redundant token spend.
Add a one-line routing hint near the top (e.g., 'New culture → Frameworks 1–3; existing culture → 4–5; intervention → 6') so Claude knows which framework to enter from for a given request.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense and opinionated with no over-explanation of concepts Claude already knows — behavioral anchors like "Decisions under €5K happen at team level, same day, no approval needed" earn their tokens. Not 5: there is trimmable redundancy — the "## Keywords" section ("culture, company culture, values, mission, vision...") duplicates frontmatter trigger terms, and "## Detailed References" repeats the two inline citations of `references/culture-playbook.md` and `templates/culture-code-template.md`. Not 3: these are minor instances, not 'some unnecessary explanation' spread through the content. | 4 / 5 |
Actionability | For an instruction-only skill, guidance is copy-paste ready: concrete workshop questions ("What would be lost if we disappeared tomorrow?"), a behavioral-anchor table with specific examples, a decision threshold table ("< 50% | Crisis | Culture emergency — all-hands intervention"), and rituals with cadences ("Weekly all-hands (30 min): company update + one win + one learning"). Not 4: no material gaps — specific examples cover the common cases across all six frameworks. | 5 / 5 |
Workflow Clarity | Frameworks are numbered in a coherent build→codify→measure→maintain lifecycle, each with internal checkpoints: the day-30 test ("How would a new hire know we actually live this on day 30?"), the promotion audit ("What did the last person who got promoted demonstrate?"), and quarterly cadence with score-band actions. Not 5: there is no explicit routing for which framework to start from given a user's request, so the top-level sequence is implicit rather than stated. Not 3: checkpoints are present within each framework, not missing or merely implicit. | 4 / 5 |
Progressive Disclosure | References are clearly signaled, described, and one level deep ("See `references/culture-playbook.md` for survey design", "See `templates/culture-code-template.md` for a complete template"), and the body appropriately keeps bulk detail out. However, scoring against the actual bundle structure: neither `references/culture-playbook.md` nor `templates/culture-code-template.md` exists in the skill directory — the bundle contains only SKILL.md — so the disclosure chain is broken and a runtime read would fail. Not 4: missing referenced files are more than a 'minor organization gap'; not 2: the inline content is well-structured and references are prominently signaled, not buried. | 3 / 5 |
Total | 16 / 20 Passed |