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.
A compact, well-structured instruction skill with concrete routing boundaries, an explicit output template, and clean workflow sequencing. Its main weaknesses are dangling cross-references buried in the output template, repeated send/CRM guardrails, and the absence of worked examples for its core artifacts.
Suggestions
Move the '../index/SKILL.md' cross-references out of the fenced output template into clearly signaled links in the body, and fix the malformed 'SKILL.md## Cross-Skill Best Practices' path.
Consolidate the 'never send / no CRM write' guardrail into one Rules entry instead of repeating it in the intro, workflow, next-steps, and rules sections.
Add one short worked example of a fit hypothesis and draft outreach line to make the output template concrete.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and assumes Claude's competence (no concept re-teaching, tight workflow steps), but 'never send / no CRM write' is repeated in four places, and strikethrough list items ('~~Sales Intelligence...') plus '{Follow the instructions...}' placeholder lines add avoidable noise. Not 5 because of this redundancy; not 3 because padding is minor rather than a pattern of unnecessary explanation. | 4 / 5 |
Actionability | Concrete, specific guidance throughout: a 4-step workflow, routing rules to three named sibling skills ('enrich-company-and-contact-data', 'find-key-internal-sources', 'plan-deal-strategy'), a full output template, and explicit next-step options. Not 5 because there are no worked examples (e.g., a sample fit hypothesis or draft message) covering the common cases; not 3 because the guidance is executable, not high-level hints. | 4 / 5 |
Workflow Clarity | The 4-step workflow is clearly sequenced with explicit decision handling ('If a material anchor is missing, ask only for the smallest input that changes the research path') and evidence/inference separation. Not 5 because validation checkpoints are lighter than the anchor's explicit validate-fix-retry loops (e.g., no step to verify sourced claims before drafting); not 3 because the sequence and guardrails are explicit, not implicit. | 4 / 5 |
Progressive Disclosure | Sections are well-organized and the skill is self-contained, but its only references ('plugins/sales/skills/index/SKILL.md', '../index/SKILL.md#limitations-and-improvements') are dangling, unverifiable, one malformed ('SKILL.md## Cross-Skill Best Practices'), and buried inside the fenced output template as curly-brace placeholders. Not 4 because navigation to supporting material is unclear and the references are not clearly signaled in the body; not 2 because the inline content is appropriately concise rather than monolithic. | 3 / 5 |
Total | 15 / 20 Passed |