Rowan Autosearch — Design skill for Claude Code
Agent-driven molecular optimization over chemical space, using Rowan for property scoring and RDKit for constraint-aware candidate design.
How to install Rowan Autosearch
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open K-Dense-AI/rowan-autosearch and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Rowan Autosearch does
Agent-driven molecular optimization over chemical space, using Rowan for property scoring and RDKit for constraint-aware candidate design.
Alternatives in Design
- Brand Guidelines — Apply Anthropic's official brand colors and typography to artifacts 94.1k ★
- Od Contribute — Open a first-contribution PR (or bug issue) on nexu-io/open-design — works for non-coders too 91.8k ★
- Export Diagram — Export a diagram-design HTML file to .svg and .png next to the source 27.2k ★
README
Rowan Autosearch
[](https://x.com/k_dense_ai) [](https://www.linkedin.com/company/k-dense-inc) [](https://www.youtube.com/@K-Dense-Inc) [](https://www.reddit.com/user/-k-dense-/)
**Agent-driven molecular property optimization.** You define a chemistry objective and a starting molecule. An AI coding agent (Claude Code, Cursor, Codex, OpenCode, Gemini, etc.) plays the role of a medicinal chemist: proposing analogs, scoring them with [Rowan](https://www.rowansci.com/) quantum/ML workflows, checking drug-likeness with RDKit, and iterating until a winning candidate emerges.
`rowan-autosearch` is the **scoring harness and lab notebook** the agent uses to do this rigorously: every candidate, every Rowan payload, every constraint check, and every design rationale is captured as append-only JSON and rendered into an auditable HTML report.
Joint work between [K-Dense](https://www.k-dense.ai) and [Rowan Scientific](https://www.rowansci.com/).
Recommended driver models: **Claude Opus 4.8**, **GPT 5.5**, or **Gemini 3.1 Pro**. Any agent that can read `AGENTS.md` and call CLI tools will work.
**Stay up to date:** Follow K-Dense on [X](https://x.com/k_dense_ai), [LinkedIn](https://www.linkedin.com/company/k-dense-inc), and [YouTube](https://www.youtube.com/@K-Dense-Inc) for release announcements, molecular optimization walkthroughs, and new Rowan workflow examples you can run with your own AI agent.
🎬 **New to running K-Dense skills with your agent?** [Getting Started with Scientific Agent Skills](https://youtu.be/ZxbnDaD_FVg) covers the agent setup this harness assumes.
Who This Is For
- Computational chemists r
Related Skills
Design Weaviate Collection
Design a Weaviate collection using Python client v4. Captures vectorizer selection, named vectors, multi-tenan
Conformance
Assess architecture conformance — ADR decision implementation, cross-decision consistency, design-principles a
Design Distill
Distill a candidate canonical from ≥5 successful same-type pages (harness component 08 · library-grower). Wrap
Morphogenetic Rl
RL controllers that decide WHEN/HOW to mutate a network's topology during training - controller design, govern
Adaptive Mentor
Dynamic learning style optimization with real-time pedagogical adaptation
Dna
Place a design in the measured design space — nearest systems, per-axis percentiles, outliers
Related Agents
Hypothesis Tester
Generates property-based tests that find edge cases humans miss. Use for pure functions, parsers, serializers,
Fuzzer
Generates property-based tests that find edge cases humans miss. Use for pure functions, parsers, serializers,
Prospector
Generates candidate data center sites from first principles across a region by constraint intersection, then w
