CtrlK
BlogDocsLog inGet started
Tessl Logo

tooluniverse-biomedical-fact-lookup

Answer biomedical FACTUAL / recall / multiple-choice questions by querying ToolUniverse database tools instead of answering from memory. Triggers on any 'which gene/drug/variant/disease/pathway/miRNA/TF...' lookup, any question phrased 'according to <database>' (DisGeNet, OMIM, MSigDB, miRDB, GTRD, MGI, Ensembl, ClinVar, ChEMBL, OpenTargets, Reactome, GtoPdb, UniProt...), and multiple-choice biology/medicine knowledge questions where one option must be verified against an authoritative source. NOT for analyzing user-supplied data files (CSV/VCF/h5ad → use the data-analysis router) and NOT for open-ended literature synthesis. Use whenever a single correct answer exists in a public biomedical database and could be looked up rather than guessed.

72

Quality

88%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

77%

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

A dense, highly actionable skill body with strong workflows and executable code, weakened only by some rhetorical padding and a monolithic single-file structure that doesn't offload detail to reference files. The lowest dimensions are conciseness and progressive_disclosure.

Suggestions

Trim motivational restatement: state the 'map the computed value back to option letters' rule once instead of repeating it across the computational-procedures intro and the interpretation section.

Move the long Python procedure blocks (ORF counting, restriction digest, gamete ratios) into a scripts/ or references/ file referenced one level deep, keeping SKILL.md as an overview + routing table.

Tighten RULE ZERO and the intro paragraph to a single directive sentence each; the hallucination risk is already implied by the 'never guess' mandate.

DimensionReasoningScore

Conciseness

Mostly efficient and packed with genuinely non-obvious content (tool names, MSigDB set-name conventions), but it repeats the 'final-answer discipline / map computed value to options' point and leans on motivational prose ('RULE ZERO', 'this is exactly what models hallucinate') that could be trimmed. Not the level-3 lean baseline because some tokens are rhetorical reinforcement rather than instruction.

2 / 3

Actionability

Highly actionable: a concrete routing table mapping question patterns to specific tools and parameters, plus complete, copy-paste-ready Python for ORF counting, restriction digests, and gamete ratios with named functions and example calls.

3 / 3

Workflow Clarity

Clear numbered sequences (Parse → Resolve → Query → Check → Answer; the 6-step 'in X but not Y' recipe) with explicit error-recovery feedback loops ('If a tool errors on a name, re-derive the name', 'If RestrictionBatch raises... re-run') and a checkpoint for 'Insufficient information' only on a genuine empty result.

3 / 3

Progressive Disclosure

No bundle files exist, so all content sits inline in one ~16KB SKILL.md. Sections are well-organized and references are not nested, but the detailed computational code blocks and per-database recipes are content that could be split into one-level-deep reference files, matching the 'content that should be separate is inline' anchor. Not level 1 because structure is genuinely clear; not level 3 because nothing is offloaded to references.

2 / 3

Total

10

/

12

Passed

Description

100%

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: concrete action, rich natural trigger terms, explicit when-guidance, and a clear NOT-for boundary that distinguishes it from sibling skills. Voice is third-person imperative throughout, so no specificity penalty applies.

DimensionReasoningScore

Specificity

Names a concrete action — 'Answer biomedical FACTUAL / recall / multiple-choice questions by querying ToolUniverse database tools' — and a distinct verification behavior ('one option must be verified against an authoritative source'), matching the multiple-concrete-actions anchor as closely as the commit-message good example does.

3 / 3

Completeness

Explicitly answers both what (answer factual biomedical questions by querying database tools) and when ('Triggers on...', 'Use whenever a single correct answer exists...'), with an explicit NOT-for boundary.

3 / 3

Trigger Term Quality

Excellent natural trigger coverage: 'which gene/drug/variant/disease/pathway/miRNA/TF...' and 'according to <database>' plus a long list of database names (DisGeNet, OMIM, MSigDB, miRDB, GTRD...) users would actually say.

3 / 3

Distinctiveness Conflict Risk

Clear niche (biomedical database lookups) with distinct triggers and an explicit exclusion ('NOT for analyzing user-supplied data files... use the data-analysis router'), making conflicts unlikely.

3 / 3

Total

12

/

12

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
mims-harvard/ToolUniverse
Reviewed

Table of Contents

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.