Conciseness

The prose sections (When to Use, Best Practices, Troubleshooting) are lean and assume competence, but the body inlines ~500 lines of full function implementations (find_primer_binding, restriction_digest, design_golden_gate, design_crispr_guides, score_guide, annotate_plasmid) even though equivalent CLI tools already exist in scripts/ — a large duplication that could be trimmed to short usage examples. Mostly efficient in tone, but the whole could be significantly tightened by leaning on the bundle.

Actionability

Every capability ships as complete, copy-paste-ready Python with concrete inputs and expected outputs (Quick Start runs standalone; each core capability includes runnable example invocations like the BamHI/EcoRI double digest). Not a 4 because the examples cover the common cases and require no assembly beyond pasting.

Workflow Clarity

The 'Typical Workflows' section lists three worked examples and Troubleshooting provides error recovery, and design_golden_gate does validate overhang compatibility inline. But the workflows are single-shot demonstration snippets with no explicit validation checkpoints (e.g., no verify-the-assembly-by-in-silico-re-digestion step after design), leaving the sequence present with checkpoints only implicit.

Progressive Disclosure

The bundle provides five purpose-built scripts (simulate_pcr.py, restriction_digest.py, golden_gate.py, design_primers.py, design_crispr.py) that duplicate the body's inline implementations, yet SKILL.md never references scripts/, references/, or assets/ anywhere — content that clearly belongs in the existing separate files is inlined and the bundle files are orphaned. Section headers within the body are decent, but the file split is broken, matching the anchor for inlined content with no functioning navigation to the bundle.

Specificity

Lists seven concrete actions — 'PCR amplicon prediction, restriction enzyme digestion, Golden Gate and Gibson assembly simulation, primer design, CRISPR sgRNA design, and plasmid annotation' — which comprehensively cover the molecular-cloning domain. Not a 4 because there is no meaningful gap in capability coverage; every listed action is specific and executable-sounding.

Completeness

The 'what' is clearly and comprehensively stated, but there is no 'Use when...' clause or equivalent explicit trigger guidance; the final sentence ('For protein-level sequence analysis use biopython or esm; for database lookups use gene-database or ensembl-database') is a redirect to other skills, not a statement of when to use this one. Per the rubric guideline, a missing 'Use when...' clause caps completeness at 3.

Trigger Term Quality

Strong natural terms users would actually say: 'PCR', 'primer design', 'Golden Gate', 'Gibson assembly', 'CRISPR sgRNA', 'plasmid', 'restriction enzyme digestion'. Not a 5 because a few natural variations and file formats are missing (e.g., 'digest' alone, 'guide RNA', 'cloning map', '.gb'/GenBank), leaving keyword coverage good but not fully comprehensive.

Distinctiveness Conflict Risk

Clear niche with distinctive domain triggers (Golden Gate, Gibson, sgRNA design) and explicit redirects that sharpen boundaries against 'biopython', 'esm', 'gene-database', and 'ensembl-database'. Not a 5 because this skill is built on Biopython, so a request phrased around 'biopython' for cloning tasks carries minor overlap risk with the biopython skill itself.