tooluniverse-protein-structure-prediction

Protein 3D structure prediction from sequence — ESMFold de novo prediction, AlphaFold database retrieval, experimental structures from RCSB, ProtVar variant impact assessment, ProtParam sequence properties. Use for structure prediction when no experimental structure exists, fold-confidence scoring, and structure-guided variant interpretation.

Quality

85%

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

81%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.

The body is a well-structured, actionable workflow with strong sequencing and validation checkpoints. The main weaknesses are minor verbosity in the comparison/database sections and the absence of literal executable code, which is appropriate for a tool-invocation skill but leaves actionability just below maximum.

Suggestions

Tighten or remove the 'ESMFold vs AlphaFold' prose and 'Databases Integrated' table where the same facts already appear in the phase sections, to reduce token cost.

Add one or two concrete Bash/Python snippets showing how to parse the returned pLDDT array and compute mean pLDDT, since Phase 2 instructs this computation but gives no executable code.

Consider moving the 'Limitations' and 'Databases Integrated' reference tables to a short reference section or trimming them, so the core workflow scan remains uncluttered.

Dimension	Reasoning	Score
Conciseness	Largely efficient use of tables and tight bullet workflows, with only minor over-explanation (e.g. the ESMFold-vs-AlphaFold prose and the 'Databases Integrated' table partly restate information already conveyed elsewhere).	4 / 5
Actionability	Provides concrete, executable guidance via named tools, exact parameter names, value formats (e.g. '"P04637 R175H"'), thresholds, and a parameter reference table; falls just short of fully copy-paste-ready because it relies on tool-call semantics rather than literal code/commands.	4 / 5
Workflow Clarity	A clearly sequenced six-phase workflow with explicit validation gates (pLDDT/pTM thresholds, 'do not interpret low-confidence regions'), a fallback-strategies table acting as error-recovery feedback loops, and a 'Quality Minimums' checklist for the final report.	5 / 5
Progressive Disclosure	Well-organized with clear section headers and no nested references (no bundle files exist, all detail is ToolUniverse calls); a few sections (Databases Integrated, Limitations) are inline content that adds length without aiding navigation, keeping it just below the ideal split.	4 / 5
	Total	17 / 20 Passed

Description

88%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.

The description is strong: it enumerates concrete capabilities and pairs them with explicit 'Use for' trigger guidance. Minor gains are possible by adding a few natural synonyms users might utter and ensuring the boundary with the sibling retrieval skill is hinted at directly in the description.

Dimension	Reasoning	Score
Specificity	Lists five concrete, distinct actions — 'ESMFold de novo prediction', 'AlphaFold database retrieval', 'experimental structures from RCSB', 'ProtVar variant impact assessment', 'ProtParam sequence properties' — giving comprehensive coverage of the skill's capabilities.	5 / 5
Completeness	Explicitly answers 'what' with five enumerated capabilities and 'when' with a concrete 'Use for structure prediction when no experimental structure exists, fold-confidence scoring, and structure-guided variant interpretation' clause.	5 / 5
Trigger Term Quality	Includes natural user-facing phrases ('structure prediction', 'no experimental structure exists', 'fold-confidence scoring', 'structure-guided variant interpretation'), but lacks common synonyms or shorthand a user might say (e.g. 'protein folding', 'model a protein'), keeping it just below comprehensive.	4 / 5
Distinctiveness Conflict Risk	A clear, specific niche (protein structure prediction) with distinct triggers and minimal conflict risk; only minor overlap with the sibling retrieval-only skill, which the body (not the description) disambiguates.	4 / 5
	Total	18 / 20 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: mims-harvard/ToolUniverse
Path: plugin/skills/tooluniverse-protein-structure-prediction/SKILL.md
Commit: 1b02920

Reviewed: 1 day 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.