ncbi-sequence-fetch

Retrieve protein and nucleotide sequences from NCBI databases using E-utilities. Supports direct accession lookup, CDS translation, gene+organism search, locus lookup, PubMed-linked sequences, patent protein extraction, and organism+length fallback search. Use when you need to fetch biological sequences by accession, gene name, locus tag, PubMed ID, or patent number.

Quality

96%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

92%Weight 40%Scale 1-3

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is a well-organized, action-dense reference: every subcommand has executable examples, a clear retrieval cascade sequences the workflows, and selection guidance is spelled out. The main gap is progressive disclosure — the ten subcommand sections are all inline and the bundled citation.bib is never referenced.

Suggestions

Move detailed per-subcommand usage into a separate reference file (e.g. references/SUBCOMMANDS.md) and keep SKILL.md as a concise overview with one-level-deep links, to improve progressive disclosure.

Either reference references/citation.bib from the body or remove it from the bundle, so no shipped reference file is orphaned.

Fix the patent-search examples so they render as fenced ```bash blocks like the other nine sections, restoring copy-paste-ready actionability.

Dimension	Reasoning	Score
Conciseness	The body is dominated by lean, executable bash examples and one-line subcommand descriptions, with no padding explaining concepts Claude already knows (e.g., what NCBI or FASTA is); not level 2 because there is little unnecessary explanation, though the API-key guidance is restated across Prerequisites and Core Rules.	3 / 3
Actionability	Each subcommand ships copy-paste-ready `uv run scripts/ncbi_fetch.py ...` commands with real arguments and flags, plus a clear retrieval cascade; not level 2 because the guidance is concrete and executable rather than pseudocode, though the patent-search examples lose their fenced-block formatting ("`bash uv run ...`").	3 / 3
Workflow Clarity	The "Standard Sequence Retrieval Cascade" lays out a 7-step priority order and "Interpreting Results" gives selection rules, with the cds-translate `is_genomic: true` flag acting as a checkpoint; not capped at 2 because these are read-only fetches (not destructive/batch ops) so the rubric's feedback-loop requirement does not bind.	3 / 3
Progressive Disclosure	Structure and section navigation are clear and `scripts/ncbi_fetch.py` is correctly referenced one level deep, but all ten subcommand details live inline in SKILL.md rather than being split into reference files, and the bundled `references/citation.bib` is never referenced or signaled; not level 3 because content is not appropriately split and there is an orphaned bundle file, not level 1 because the overview does serve as a usable index.	2 / 3
	Total	11 / 12 Passed

Description

100%Weight 40%Scale 1-3

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is third-person, concise, and specific, enumerating seven concrete capabilities and providing an explicit "Use when" trigger with natural domain terms. It clearly communicates both what the skill does and when to use it with minimal risk of conflicting with other skills.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete capabilities — "direct accession lookup, CDS translation, gene+organism search, locus lookup, PubMed-linked sequences, patent protein extraction, and organism+length fallback search" — matching the anchor for several specific concrete actions; not level 2 because coverage is comprehensive rather than partial.	3 / 3
Completeness	Explicitly answers both what (retrieve protein/nucleotide sequences via E-utilities with the listed capabilities) and when ("Use when you need to fetch biological sequences by accession, gene name, locus tag, PubMed ID, or patent number"); not level 2 because the trigger is explicit, not merely implied.	3 / 3
Trigger Term Quality	The "Use when" clause surfaces natural domain terms a user would say — "accession, gene name, locus tag, PubMed ID, or patent number" — giving good coverage; not level 2 because it lists several common variations rather than just one or two.	3 / 3
Distinctiveness Conflict Risk	Clearly scoped to "NCBI databases using E-utilities" with distinct triggers, occupying a clear niche unlikely to trigger the wrong skill; not level 2 because the NCBI/E-utilities framing plus enumerated input types make it sharply distinguishable.	3 / 3
	Total	12 / 12 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: google-deepmind/science-skills
Path: skills/ncbi_sequence_fetch/SKILL.md
Commit: 0b42509

Reviewed: 8 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.