alphafold-db

Access over 200M protein structures from AlphaFold DB; use when you need to retrieve predicted 3D structures (PDB/mmCIF), confidence metrics (pLDDT/PAE), or protein metadata by UniProt accession.

Quality

89%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

80%Weight 40%Scale 1-5

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

A well-structured, highly actionable simple skill: the CLI example matches the bundled script exactly and API details are correctly pushed to a real one-level reference. The main gap is the absence of any validation/verification guidance despite the skill advertising batch processing, which caps workflow clarity.

Suggestions

Add a verification step for batch runs, e.g. after invoking the script, check that <UNIPROT_ID>.<cif|pdb> and <UNIPROT_ID>_metadata.json exist and are non-empty, and note that the script exits nonzero on failed fetches so batch loops can skip and report failures.

Trim redundancy: merge the overlapping 'When to Use' and 'Key Features' bullets, and consolidate the hedged metadata phrasing ('includes confidence/URL fields such as pLDDT-related information') into one precise statement.

State the PAE download URL's availability explicitly (the script captures paeDocUrl in metadata but never downloads the PAE file), so users know whether PAE data retrieval requires an extra step.

Dimension	Reasoning	Score
Conciseness	The body is mostly lean — one copy-paste command plus exact artifact names — but has minor trims available: 'Key Features' repeats 'When to Use' content, and hedged phrasing like '(includes confidence/URL fields such as pLDDT-related information)' appears in both Example Usage and Implementation Details.	4 / 5
Actionability	The example commands ('python scripts/fetch_structure.py --uniprot_id P00520 --output_dir ./out --format cif') are copy-paste ready, match the actual script's argparse interface (verified in scripts/fetch_structure.py: --uniprot_id, --output_dir, --format cif\|pdb, default cif), cover both formats, and name the expected output files exactly.	5 / 5
Workflow Clarity	The single action is unambiguous, but the skill explicitly targets batch processing ('Simple CLI workflow suitable for scripting and batch processing', 'fetches structures + confidence metrics for many proteins') and provides no validation/verification step — only a descriptive 'Expected outputs' listing, leaving checkpoints implicit. Per the rubric's batch-operation cap, workflow clarity cannot exceed 3 without validation.	3 / 5
Progressive Disclosure	Under 50 lines with well-organized sections; API endpoint details are appropriately split into references/api_reference.md (verified: real file, one level deep, clearly signaled under Implementation Details) rather than inlined, and the single action plus outputs are fully described in the body.	5 / 5
	Total	17 / 20 Passed

Description

95%Weight 40%Scale 1-5

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

A strong description: explicit what-and-when structure, comprehensive natural trigger terms for its niche, and negligible conflict risk. The only flaw is the second-person voice ('use when you need'), which the rubric penalizes on specificity.

Dimension	Reasoning	Score
Specificity	The description lists multiple concrete actions with format qualifiers ('retrieve predicted 3D structures (PDB/mmCIF), confidence metrics (pLDDT/PAE), or protein metadata by UniProt accession'), matching the comprehensive-coverage 5 anchor. However, the second-person phrasing 'use when you need to retrieve' violates the third-person voice requirement, reducing the score by 1.	4 / 5
Completeness	It explicitly answers both what ('Access over 200M protein structures from AlphaFold DB') and when ('use when you need to retrieve predicted 3D structures (PDB/mmCIF), confidence metrics (pLDDT/PAE), or protein metadata by UniProt accession') with concrete trigger phrases, mirroring the 5 anchor example.	5 / 5
Trigger Term Quality	Comprehensive coverage of the natural domain vocabulary users would actually say: 'protein structures', 'AlphaFold DB', 'PDB/mmCIF', 'pLDDT', 'PAE', and 'UniProt accession', including file format/extension terms as the 5 anchor requires.	5 / 5
Distinctiveness Conflict Risk	A clear niche (AlphaFold DB) with distinct domain triggers (pLDDT, PAE, UniProt) that are unlikely to fire for any other skill; minimal conflict risk.	5 / 5
	Total	19 / 20 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: aipoch/medical-research-skills
Path: scientific-skills/Evidence Insight/alphafold-db/SKILL.md
Commit: 686e09d

Reviewed: 12 hours 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.