tooluniverse-toxicology

Drug and chemical toxicity assessment via adverse outcome pathways (AOPs), real-world FAERS adverse event signals, FDA labels, and toxicogenomic associations. Triangulates molecular initiating event to cellular outcome to organ-level toxicity to clinical adverse event. Use for hepatotoxicity/cardiotoxicity/nephrotoxicity prediction and toxicology reports.

Quality

81%

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

71%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-organized, domain-rich skill body with concrete tool signatures, thresholds, fallbacks, and a clear phased workflow. It loses points on progressive disclosure (everything inlined, no reference files) and on workflow clarity (validation checkpoints are implicit) despite an explicit 'COMPUTE, DON'T DESCRIBE' directive that lacks a sample executable script.

Suggestions

Add one concrete, copy-pasteable Python example (e.g., a pandas snippet computing PRR/ROR from FAERS tool output) to satisfy the 'COMPUTE, DON'T DESCRIBE' directive and lift actionability.

Insert explicit validate-then-proceed checkpoints between phases (e.g., 'Only proceed to Phase 1 once PubChem CID/SMILES is confirmed' and 'Abort Phase 2 if no reaction has >=3 cases') to add feedback loops.

Move the bulk tool parameter reference and signal-threshold tables into a references/ file (e.g., TOOL_REFERENCE.md) and link from SKILL.md to improve progressive disclosure.

Dimension	Reasoning	Score
Conciseness	Largely lean and task-focused with tight tables and tool signatures, but the Domain Reasoning paragraph and some explanatory notes (e.g., the temporal acute/chronic aside) go slightly beyond what Claude already knows about toxicology.	4 / 5
Actionability	Provides concrete tool names with exact parameter signatures, signal-threshold tables, and a parameter-reference table flagging common wrong/correct args; minor gap is the absence of an executable Python example despite the 'COMPUTE, DON'T DESCRIBE' directive.	4 / 5
Workflow Clarity	Clear phased sequence (0–4 + Synthesis) with decision logic and fallback chains, plus report-first progressive updates; falls short of 5 because validation/verification checkpoints between phases (e.g., confirming disambiguation succeeded before querying) are mostly implicit rather than explicit validate-then-proceed gates.	4 / 5
Progressive Disclosure	No bundle files exist and all content is inlined into a single SKILL.md; the tool reference, parameter table, and risk-classification detail could appropriately live in separate reference files, so structure is reasonable but not split across files as the anchor for 4–5 expects.	3 / 5
	Total	15 / 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.

A strong, specific description that clearly states both the capability and when to use it, with concrete domain terminology and low conflict risk. Minor improvement possible by adding lay synonyms like 'drug safety' or 'side effects' for broader trigger coverage.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete capabilities (AOP-based assessment, FAERS signal detection, FDA label mining, toxicogenomic associations) and specifies the triangulation across MIE → cellular → organ → clinical, giving comprehensive coverage.	5 / 5
Completeness	Explicitly answers both 'what' (toxicity assessment via AOP/FAERS/FDA/CTD triangulation) and 'when' ('Use for hepatotoxicity/cardiotoxicity/nephrotoxicity prediction and toxicology reports'), with concrete trigger phrases.	5 / 5
Trigger Term Quality	Includes strong natural terms users would say ('hepatotoxicity/cardiotoxicity/nephrotoxicity prediction', 'toxicology reports', 'toxicity assessment') plus domain keywords (AOPs, FAERS), but lacks common lay synonyms like 'drug safety' or 'side effects'.	4 / 5
Distinctiveness Conflict Risk	Occupies a clearly distinct niche (toxicology via AOPs and FAERS signal detection) with specific tool/data sources named, making overlap with other skills unlikely.	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: plugins/tooluniverse/skills/tooluniverse-toxicology/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.