jincinga24-hue

Chem Agent — AI skill for Claude Code

AI community

Autonomous chemical engineering agent: LLM tool-use over RDKit, Antoine thermo, Python+scipy, arxiv literature.

How to install Chem Agent

This entry records only its repository, not the path inside it, so there is no exact command to give. Open jincinga24-hue/chem-agent and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Chem Agent does

Autonomous chemical engineering agent: LLM tool-use over RDKit, Antoine thermo, Python+scipy, arxiv literature. 23-problem benchmark across 10 ChemE subdomains.

Alternatives in AI

WeKnora — Open-source LLM knowledge platform: turn raw documents into a queryable RAG, an autonomous reasoning agent, an 26.9k ★
Welcome — AI Research Skills — You now have access to 86 production-ready skills covering the entire AI research lifecycle: literature survey 5.4k ★
Codeflow — OpenVibeCoding is the command tower for AI engineering: plan, delegate, track, resume, and prove long-running 206 ★

README

ChemAgent

Autonomous chemistry/chemical-engineering research agent. Plans, reasons, and solves ChemE problems using LLM tool-use over molecular, thermodynamic, computational, and literature tools.

**Backend:** [`claude -p`](https://docs.claude.com/en/docs/claude-code/cli-reference) — uses your Claude Code auth. No separate API key required.

**v0.2 benchmark:** 228/230 (99.1%), 22/22 numerical problems correct across 10 ChemE subdomains.

**v0.3 (in progress):** added polymer chemistry (RAFT kinetics), peptide / AMP descriptors, structured trace logging, and RAG over a polymer-chemistry corpus — 29 benchmark problems across 17 categories, **104 unit tests**, 96.9% on the v0.3 baseline run.

Why

Most LLM agents are built by CS engineers against software tasks. This one is built by a chemical engineering student to tackle *domain* problems — unit operations, kinetics, thermodynamics, separations — the kind of work a process engineer or research chemist does. It's positioned at the intersection of AI agent building and chemical engineering, a combination that's scarce in 2026.

What it does

Given a problem like:

Design a CSTR for aspirin production via salicylic acid + acetic anhydride with rate r = k·[SA]·[AA], k = 0.001 L/(mol·s). Feed [SA]_0 = 2.0 mol/L, [AA]_0 = 2.2 mol/L (10% excess). Target 80% conversion of SA. Plant must produce 100 kg/day of aspirin (MW 180). Compute the required CSTR volume.

ChemAgent:

Plans a solution path (ReAct loop, JSON actions)
Calls tools — molecular lookup (RDKit), vapor pressure (Antoine), Python execution with scipy, arxiv literature search
Returns a reasoned answer with equations, units, and final value
Is scored by an LLM-judge against a ground-truth benchmark