Conciseness

Lean and table-driven throughout with no conceptual padding — no explanation of what meta-analysis or these statistics are; every rule carries a specific operational payload ("2 × 1.96 assumes a 95% CI; use 2 × 1.645 for 90%", "Q_p_value < 0.10 → statistically significant heterogeneity", "With ≥10 studies, inspect a funnel plot / Egger's test; with <10, state that small-study bias cannot be assessed"). Matches the anchor where every token earns its place; even the limitations section is behavior-directing rather than explanatory.

Actionability

Fully executable guidance: a copy-paste `tu run MetaAnalysis_run` command with real values and a documented return schema, the helper script invocation with its exact CSV column sets (verified against scripts/meta_analysis.py), exact conversion formulas for each input form, and a ready-to-use reporting template with a worked example ("pooled OR was 1.51 (95% CI 1.33–1.72 ...)"). Covers the common cases across all effect-measure types.

Workflow Clarity

A five-step numbered workflow with the error-prone step explicitly flagged ("Step 1 ← do this carefully", "THE ERROR-PRONE STEP"), input validation rules (log-transform ratio measures, "One direction", "Same effect measure"), decision tables for fixed-vs-random and I² interpretation, a scale foot-gun warning with explicit remediation ("exp() the pooled_effect and CI bounds back"), conditional guidance for edge cases (borderline I², I²>75% → don't pool), and an ordered reporting checklist. Since this is an analytical (non-destructive) workflow, the batch/destructive validation cap does not apply; the checkpoints present match the top anchor.

Progressive Disclosure

SKILL.md is an appropriately sized operational overview; the single bundle file (scripts/meta_analysis.py) exists, is clearly signaled with its usage and supported column sets, and all references are one level deep — including the cross-skill pointer to ../tooluniverse-statistical-modeling/references/cox_regression.md and the related-skills section. Nothing that belongs in a separate reference file is inlined (the conversion and interpretation tables are needed on every run), and navigation via the step headers is easy.

Specificity

Lists multiple concrete actions — "pool effect sizes across studies ... with fixed- or random-effects models", "quantify heterogeneity (Q, I², τ²)", "build a forest plot", and converting "OR/HR/CI, two-group means±SD, proportions, and correlations into the (effect, SE)" — with comprehensive coverage across measure types (odds ratios, risk ratios, hazard ratios, mean differences, correlations, GWAS betas). This matches the anchor for multiple specific concrete actions with comprehensive coverage; nothing is vague or generic.

Completeness

Explicitly answers both questions: the "what" (pooling with fixed/random-effects models, heterogeneity quantification, forest plot, effect-size conversion) and a concrete "Use when you have results from MULTIPLE studies and need a single pooled estimate, or to synthesize evidence from a systematic review / multiple GWAS / replicated experiments" trigger clause. Matches the anchor for clearly and explicitly answering both what AND when with concrete trigger phrases.

Trigger Term Quality

Natural phrases users would actually say are comprehensively covered with synonyms: "meta-analysis / evidence synthesis", "pooled estimate", "systematic review", "forest plot", "GWAS", "replicated experiments", plus the specific effect-measure names. Clearly matches the top anchor for comprehensive natural-term coverage rather than anchor 4, which is for descriptions with notable missing synonyms.

Distinctiveness Conflict Risk

"Meta-analysis / evidence synthesis" gated on "results from MULTIPLE studies" is a clear niche with distinct triggers (pooling, forest plot, I²), but terms like "correlations", "GWAS betas", and "mean differences" create minor overlap risk with closely related sibling statistical/GWAS skills. Anchor 4 ("mostly distinct; minor overlap risk with closely related skills") fits slightly better than anchor 5 because the description does not draw an explicit boundary against those adjacent single-study-analysis skills.