Content
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a strong, highly actionable screening skill with clear workflows, concrete scoring criteria, and well-structured output templates. Its main weakness is moderate verbosity — the version notes, YC-specific methodology, and some repeated guidance could be trimmed or moved to reference files. The progressive disclosure structure is reasonable but would benefit from the referenced bundle files actually existing.
Suggestions
Move the YC Batch Evaluation section to a separate reference file (e.g., references/yc-methodology.md) and link to it, keeping the main skill focused on the universal screening workflow.
Remove or minimize the v2.0 changelog at the top — embed the behavioral changes directly into the relevant sections rather than listing them separately.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Generally efficient and well-structured with tables and concise formatting, but includes some unnecessary context like the version change notes and the Guillermo Toro methodology section adds weight. The guidelines table repeats information already stated elsewhere (e.g., 'No false eliminators' restates the hard filters section). Overall mostly lean but could be tightened. | 2 / 3 |
Actionability | Highly actionable with concrete scoring criteria, specific thresholds (e.g., >=5 signals = Tier 1), exact search query templates, a worked scoring example ('Acme AI'), and a precise output template with field-level formatting. The signal examples are specific and quantified (>$500K ARR, 100+ customers, >20% MoM growth). | 3 / 3 |
Workflow Clarity | Clear four-phase workflow with time estimates per phase, explicit validation in Phase 3 for Tier 1/2 companies, a completion checklist with verification steps, and a force-prioritization guardrail (re-tighten if >15). The feedback loop of 'flag for Phase 3 validation' in the scoring example demonstrates error recovery thinking. | 3 / 3 |
Progressive Disclosure | References to external files are well-signaled (SCREENING-CRITERIA.md, screening-worksheet.md, shortlist-template.md) and one level deep, but no bundle files are provided to verify these exist. The YC batch evaluation section and guidelines table could arguably be split into reference files to keep the main skill leaner. The main body is somewhat long for an overview document. | 2 / 3 |
Total | 10 / 12 Passed |