tooluniverse-protein-sae-variant-interpretation

Interpret a missense variant via ESMC-6B Sparse Autoencoder (SAE) feature activations. For a given protein + variant, computes which interpretable SAE features (catalytic, ligand-binding, PTM, structural motif, domain, etc.) are lost or gained at the mutation site. Use when standard pathogenicity scores (AlphaMissense, ClinVar) say a variant is damaging but you need a MECHANISTIC explanation — e.g. 'why is this variant LoF?' Complements (does not replace) variant-interpretation and variant-to-mechanism skills, which focus on ACMG classification or regulatory mechanism.

Quality

87%

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

82%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 highly actionable with executable code throughout and a clear sequenced workflow backed by honest limitations and a cross-validation pattern. Minor tightening of conceptual asides and an explicit checkpoint on the batch path would lift it further.

Suggestions

Trim restatements of general concepts Claude already knows (e.g. the opening definition of SAE features) to improve token efficiency.

Add an explicit validation/verification checkpoint after the batch saturation call before reporting results.

Consider moving the 6-category interpretation table into a referenced file to keep SKILL.md a tighter overview, if a references/ bundle is added.

Dimension	Reasoning	Score
Conciseness	Mostly lean with direct tool calls and minimal padding, but a few explanatory asides (e.g. 'SAE features are interpretable latent dimensions of the model's hidden state — many activate on biologically meaningful patterns') restate what Claude largely already knows.	4 / 5
Actionability	Fully executable Python tool calls with concrete parameters cover the common cases — quick composite path, raw delta, batch saturation, and the long per-residue path — making it copy-paste ready.	5 / 5
Workflow Clarity	A clear 5-step sequence with explicit error handling (Step 3 residue-mismatch returns an explicit error; ref_aa mismatch in tools), but the batch saturation operation (20 Forge calls) lacks an explicit verify-checkpoint, leaving a minor validation gap.	4 / 5
Progressive Disclosure	No bundle files exist; the single file is well-sectioned (When to use, Required inputs, Workflow, Interpretation table, Limitations, Cross-validation, Reporting) with easy navigation and no nested references, though some sizable reference material (e.g. the 6-category interpretation table) is inlined rather than split out.	4 / 5
	Total	17 / 20 Passed

Description

92%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 specific, complete, and well-differentiated, with strong trigger terms and explicit 'Use when' guidance plus boundary notes against sibling skills. It is an exemplar of a clear, well-scoped skill description.

Dimension	Reasoning	Score
Specificity	Names the domain and multiple concrete actions — 'Interpret a missense variant via ESMC-6B Sparse Autoencoder (SAE) feature activations' and 'computes which interpretable SAE features… are lost or gained at the mutation site' — giving comprehensive, specific coverage.	5 / 5
Completeness	Explicitly answers both what ('computes which…SAE features…are lost or gained') and when ('Use when standard pathogenicity scores…say a variant is damaging but you need a MECHANISTIC explanation') with concrete trigger phrases.	5 / 5
Trigger Term Quality	Strong natural terms practitioners actually say ('missense variant', 'damaging', 'AlphaMissense', 'ClinVar', 'why is this variant LoF?'), though a few synonyms/common phrasings are absent, keeping it just below comprehensive.	4 / 5
Distinctiveness Conflict Risk	A clear niche (ESMC-6B SAE feature interpretation) with explicit boundary guidance — 'Complements (does not replace) variant-interpretation and variant-to-mechanism skills' — minimizing conflict risk.	5 / 5
	Total	19 / 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-sae-variant-interpretation/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.