pdb-database

Use when you want to search for or download experimentally-determined 3D structures for biomolecules (proteins, nucleic acids, bound ligands). Supports searching by sequence similarity, structure similarity, chemical and other attributes. Also use to get metadata about biomolecular structure experiments.

Quality

90%

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

80%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 highly actionable and token-efficient, with concrete executable commands and real example queries for each workflow and clear step sequencing. Its main weaknesses are implicit (rather than explicit) validation checkpoints in the workflows and a somewhat monolithic structure that leaves a bundled reference file orphaned.

Suggestions

Add explicit validation/verification checkpoints to the workflows (e.g. after a search, verify result count with --count_only before fetching full results; after download, confirm expected file count/format) to support validate-then-proceed feedback loops for batch operations.

Reference references/citation.bib from the body (e.g. in a Citations section) or remove it if unused, so the bundled reference is discoverable rather than orphaned.

Consider moving the large inline JSON/GraphQL example-query blocks into a separate reference file (e.g. references/example_queries.md) linked from each workflow to reduce body length and improve navigation.

Dimension	Reasoning	Score
Conciseness	The body is lean and assumes Claude's competence — direct script invocations and copy-paste example queries with no padding about what proteins or PDB files are; the 'basic PDB concepts' section covers domain-specific schema notions (entity vs instance, label vs auth numbering) rather than general knowledge, so tokens earn their place.	3 / 3
Actionability	Fully executable commands throughout (e.g. 'uv run scripts/fetch_schema.py --api search_structure --output schema_structure.txt', 'uv run scripts/search_pdb.py --query ... --output results.json') plus multiple complete real JSON/GraphQL example queries, matching the 'fully executable, copy-paste ready' anchor.	3 / 3
Workflow Clarity	Multi-step processes are clearly numbered (fetch schema → grep → compose/run for both attribute and metadata workflows), but validation checkpoints are implicit rather than explicit validate-then-proceed loops, and batch download lacks an explicit verification step, fitting the score-2 'sequence present but checkpoints missing or implicit' anchor.	2 / 3
Progressive Disclosure	Sections are well organized and the four scripts are clearly signaled one-level-deep references, but the lengthy inline query-example blocks are content that could be split into separate reference files and the bundled references/citation.bib is never referenced from the body, fitting the score-2 'some structure, content that should be separate is inline' anchor.	2 / 3
	Total	10 / 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 strong: it states concrete capabilities, includes explicit 'Use when' triggers covering both what and when, and carves out a distinct niche unlikely to conflict with other skills. It is concise without padding or over-claims.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'search for or download experimentally-determined 3D structures', 'searching by sequence similarity, structure similarity, chemical and other attributes', and 'get metadata about biomolecular structure experiments' — matching the 'lists multiple specific concrete actions' anchor rather than the score-2 partial coverage.	3 / 3
Completeness	Explicitly answers both what (search/download structures, get metadata) and when via two 'Use when...'/'Also use to...' trigger clauses, matching the score-3 anchor; not score-2 because the when-guidance is explicit rather than implied.	3 / 3
Trigger Term Quality	Natural phrasings a user would say appear throughout ('search for... 3D structures', 'download... structures', 'proteins, nucleic acids, bound ligands', 'sequence similarity', 'metadata about biomolecular structure experiments'); above the score-2 anchor which expects only some relevant keywords with missing variations.	3 / 3
Distinctiveness Conflict Risk	Occupies a clear niche (biomolecular 3D structures with sequence/structure/chemical similarity search) with distinct triggers unlikely to overlap other skills, matching the 'clear niche with distinct triggers' anchor.	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/pdb_database/SKILL.md
Commit: 0b42509

Reviewed: 7 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.