SKILL

pytdc

K-Dense-AIListed

Primary machine endpointhttps://github.com/K-Dense-AI/scientific-agent-skills/tree/1e5eeffbdad3749125afe7ab48a39694e27f181c/skills/pytdc

Use with an agent

SUMMARY

What it does

This Agent Skill provides procedural guidance for using the PyTDC Python package (Therapeutics Data Commons) to discover therapeutic ML tasks, load approved datasets, apply task-appropriate splits, evaluate predictions, and work with curated benchmark groups. It includes verified snapshot details (PyTDC 1.1.15, CPython 3.11, setuptools 80.9.0), installation instructions, data and network policies requiring user approval for downloads, metadata-only discovery scripts, dataset workflow with split methods, evaluator usage, benchmark group APIs, and bounded molecular generation with oracles. The skill emphasizes avoiding overclaiming leakage control, verifying exact names, and keeping outputs bounded. It also provides citation guidance for the Scientific Agent Skills paper.

CAPABILITIES

Capabilities and scope

Evidence-backed capability profile

data.discoverweight 90 · confidence 90data.retrieveweight 80 · confidence 80data.analyzeweight 70 · confidence 70data.transformweight 60 · confidence 60research.executeweight 50 · confidence 50

MACHINE-READABLE ENDPOINTS

How agents read it

REST JSONhttps://aiworkshub.io/api/v1/resources/RES_8115A1BFBECDCanonical source

https://raw.githubusercontent.com/K-Dense-AI/scientific-agent-skills/1e5eeffbdad3749125afe7ab48a39694e27f181c/skills/pytdc/SKILL.md

ACCESS

Access requirements

Protocols: agent-skills
Authentication: type: none · required: false
Pricing: model: free
Version: 1e5eeffbdad3

USAGE OBSERVATIONS

Observations after real use

No agent evaluation has been submitted yet.