Content
31%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 padded CTO playbook: mostly generic leadership content Claude already knows, with pseudocode bash blocks and references to script/reference files that are absent from the bundle. The few genuinely valuable parts (ratios, DORA targets, evaluation steps) are buried in filler. It needs drastic pruning and either real reference files or removal of the broken pointers.
Suggestions
Cut generic leadership filler Claude already knows (Weekly Cadence, Crisis Management, Stakeholder Management, Communication Templates, Tools & Resources book/community lists) — keep only skill-specific guidance such as the ratios, DORA targets, and ADR workflow, cutting the body to a fraction of its size.
Ship the referenced bundle files (references/architecture_decision_records.md, references/technology_evaluation_framework.md, references/engineering_metrics.md, scripts/tech_debt_analyzer.py, scripts/team_scaling_calculator.py) or delete the pointers — every reference and script command in the body currently points to a nonexistent file, so `python scripts/tech_debt_analyzer.py` fails on invocation.
Move detailed playbooks (crisis response, quarterly planning, stakeholder reporting) into one-level-deep reference files with clearly signaled links, and convert the ```bash blocks into either real executable commands or plain markdown lists, since they currently mix shell syntax with prose bullets that cannot run.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~400-line body is heavily padded with generic leadership knowledge Claude already has: a day-by-day "Weekly Cadence", "Crisis Management" basics ("Assess severity, Activate incident team"), "Stakeholder Management", communication templates, and a Tools/Resources catalog ("Slack, Zoom, Loom", book recommendations, communities). This matches the anchor-1 pattern of extensive explanation of known concepts; it is not a 2, where only a few sections would be padded — here the majority of the document is non-skill-specific filler. | 1 / 5 |
Actionability | There is some concrete guidance — specific ratios ("Manager:Engineer = 1:8", "Senior:Mid:Junior = 3:4:2"), DORA targets ("Deployment Frequency: >1/day", "Change Failure Rate: <15%"), and week-numbered vendor evaluation steps — but the ```bash blocks are not executable (they mix shell commands with bullet-list prose), and the referenced scripts (scripts/tech_debt_analyzer.py, scripts/team_scaling_calculator.py) do not exist in the bundle. This lands on anchor 3 (some concrete guidance but incomplete, pseudocode instead of executable code, missing key details) rather than 4, which requires mostly executable guidance with only minor gaps. | 3 / 5 |
Workflow Clarity | Multi-step sequences are listed (incident response with time windows, vendor evaluation over 4 weeks, 5-step cloud migration), so a sequence exists, but there are no validation checkpoints, decision criteria, or error-recovery feedback loops anywhere. This matches anchor 3 ("steps listed but validation gaps; sequence present but checkpoints missing or implicit"); it is above anchor 2 because the sequences are concrete and time-boxed, and below anchor 4 because every workflow lacks any verification step. | 3 / 5 |
Progressive Disclosure | Per the guideline to score against the actual bundle structure, all five referenced files (references/architecture_decision_records.md, references/technology_evaluation_framework.md, references/engineering_metrics.md, scripts/tech_debt_analyzer.py, scripts/team_scaling_calculator.py) are missing — the pointers are broken. Meanwhile ~350 lines of content that clearly belongs in separate reference files (Weekly Cadence, Crisis Management, Stakeholder Management, Communication Templates) are inlined in SKILL.md. This matches anchor 2 (content that clearly belongs in separate files is inlined; references broken/buried); the Quick Start's pointer style keeps it from anchor 1's monolithic wall of text. | 2 / 5 |
Total | 9 / 20 Passed |