tooluniverse-residue-functional-mechanism-interpretation

Given a set of residues in a protein, explain WHY they are functionally critical by combining structural context (binding interface, ligand pocket, core, secondary structure), UniProt features (active sites, binding sites, PTM sites, disulfides), optional SAE feature evidence, and optional DMS data. Accepts residues from any source: DMS hotspots (top-K by max effect), ClinVar recurrent variants, literature-reported hot regions, evolutionarily conserved positions, or user-curated lists. Returns a per-cluster mechanism call: catalytic / ligand-binding / interface / structural-core / PTM / regulatory / unknown.

Quality

70%

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

Fix and improve this skill with Tessl

tessl review fix ./plugin/skills/tooluniverse-residue-functional-mechanism-interpretation/SKILL.md

Quality

Content

70%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 strong, highly actionable workflow body: real executable code, explicit validation checkpoints (premise check, alignment assert), honest limitations, and clear path contracts. The weaknesses are length (~500 lines inlined with no reference files despite the small-skills exception not applying) and a few code blocks that assume undefined variables.

Suggestions

Move the Step 7 visualization code (~75 lines) and the interpretation/cross-reference tables into a references/ file (e.g. references/visualization.md) and link them one level deep, which would improve both conciseness and progressive_disclosure.

Define or note the provenance of variables used in code snippets (`pos_index`, `ref_sequence`, `wt_vec`, `positions`, `sequence`, `amino_acid_order`, `hotspot_results`) so each block is copy-paste ready.

Trim the Step 7 boilerplate to the three cell-color rules and alignment landmark, leaving standard matplotlib mechanics to Claude, and tighten the KRAS premise-check anecdote to the mismatch-reporting rule it illustrates.

Dimension	Reasoning	Score
Conciseness	The ~500-line body is mostly dense domain-specific guidance (tool costs, pitfalls, decision rules) that Claude would not know, but it could be tightened: the Step 7 matplotlib block (~75 lines) largely spells out standard plotting Claude already knows, and the KRAS premise-check anecdote runs long.	3 / 5
Actionability	Concrete, near-executable Python throughout — hotspot detection, clustering, structural/UniProt/SAE evidence gathering, permutation testing, mechanism calling, and plotting, with real tool names and parameters. Minor gaps: several code blocks reference variables never defined in the skill (e.g. `pos_index` in Step 0, `ref_sequence`/`wt_vec` in Step 4, `positions`/`sequence`/`amino_acid_order`/`hotspot_results` in Step 7).	4 / 5
Workflow Clarity	Steps 0-7 are clearly sequenced with two explicit entry paths (A/B), an explicit premise-check decision rule with numeric thresholds and transparent mismatch reporting, a landmark alignment assertion in Step 7, and documented fallbacks (single-position clusters fall back to descriptive ranking; below-top-50% ranks are reported up front).	5 / 5
Progressive Disclosure	No bundle files exist (no references/, scripts/, or assets/), so all ~500 lines are inlined in one SKILL.md. Internal structure is good, but content that clearly belongs in separate files is inline — the Step 7 plotting code, the interpretation table, and the cross-reference roster are natural reference-file candidates for a skill this long.	3 / 5
	Total	15 / 20 Passed

Description

71%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 specific, information-dense description with comprehensive inputs and a concrete output vocabulary, written in third person. Its main weakness is the absence of an explicit 'Use when...' trigger clause, which leaves the activation conditions only implied and slightly raises conflict risk with sibling variant-interpretation skills.

Suggestions

Add an explicit 'Use when...' clause, e.g. 'Use when the user asks why specific residues or mutation hotspots are functionally important' — this would lift completeness from 3 toward 5.

Add one or two natural user phrasings ('mutation', 'hotspot', 'why do mutations at these positions keep appearing') to strengthen trigger-term coverage.

Add a one-line contrast with the single-variant sibling skills (e.g. 'For single-variant analysis use ... instead') to reduce overlap risk.

Dimension	Reasoning	Score
Specificity	The description lists multiple concrete actions and inputs comprehensively: 'combining structural context (binding interface, ligand pocket, core, secondary structure), UniProt features (active sites, binding sites, PTM sites, disulfides), optional SAE feature evidence, and optional DMS data', enumerates accepted residue sources, and specifies a concrete output vocabulary ('catalytic / ligand-binding / interface / structural-core / PTM / regulatory / unknown').	5 / 5
Completeness	The 'what' is clear and detailed, but there is no 'Use when...' clause or equivalent explicit trigger guidance — the 'when' is only weakly implied by 'Accepts residues from any source', which caps completeness at 3 per the judging guidelines.	3 / 5
Trigger Term Quality	Good domain keyword coverage — 'residues', 'functionally critical', 'DMS hotspots', 'ClinVar recurrent variants', 'conserved positions', 'mechanism' — but a few natural user phrasings are missing (e.g. 'mutation', 'why do mutations at these sites...', 'hot regions' appears only once).	4 / 5
Distinctiveness Conflict Risk	The residue-mechanism niche is well differentiated ('per-cluster mechanism call', 'Accepts residues from any source'), but trigger terms like DMS, SAE, and variants overlap with the closely related sibling skills this SKILL.md body itself names (variant-predictor-dms-validation, protein-sae-variant-interpretation, protein-lof-mechanism), creating minor overlap risk.	4 / 5
	Total	16 / 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
skill_md_line_count	SKILL.md is long (506 lines); consider splitting into references/ and linking	Warning

	Total	15 / 16 Passed

Repository: mims-harvard/ToolUniverse
Path: plugin/skills/tooluniverse-residue-functional-mechanism-interpretation/SKILL.md
Commit: 164aaf1

Reviewed: 6 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.