Protein Engineer

In-silico protein design: from a target and mapped epitope to structurally validated, developable binders and scaffolds.

A specialist agent profile and its companion skills for Claude Science. In-silico protein design bench: takes a target + epitope to designed, structurally validated, developable binders and multivalent scaffolds. Composes the binder-design, developability-hardening, and multivalent-display skills with the structure-model skills (RFdiffusion/MPNN/AlphaFold/Boltz/Chai/ESMFold) and remote GPU compute, with strict fold-back and developability discipline.

What's in this repository

agent/
  profile.json        # picker metadata (name, description, access)
  system_prompt.md    # the agent's identity / opening system prompt
skills/               # 5 skill(s), each with SKILL.md (+ kernel.py helpers)
install.py            # one-shot, idempotent installer

Skills bundled

• antigen-epitope-pipeline
• binder-design-campaign
• binder-developability-hardening
• compute-cost-audit
• multivalent-display-design

Install it in Claude Science

Easiest — paste one prompt to your Claude Science agent

Please install the Protein Engineer from this GitHub repo: https://github.com/ruthannepai-tech/protein-engineer Clone or download it, then run install.py from the repo root in the repl tool with exec(open("install.py").read()). It will create and publish the skills and create the PROTEIN_ENGINEER agent profile with full access. Then create its environment and offer to switch me to it.

When it finishes, Protein Engineer appears in your agent picker.

One command — if you already have the repo

From the repo root, in the Claude Science repl tool:

exec(open("install.py").read())

install.py publishes every skill under skills/ and creates the PROTEIN_ENGINEER agent profile with full catalog + connector access. It is idempotent — safe to re-run; it updates in place. It finishes by printing the optional environment setup and the switch command:

manage_environments(mode='create', name='protein-engineer',
    packages=['requests'], python_version='3.13')
host.agents.switch('PROTEIN_ENGINEER')

Manual

The SKILL.md / kernel.py files and agent/ files are plain text and fully define everything; create each skill and an unrestricted profile named PROTEIN_ENGINEER by hand if you prefer.

A note on shared skills

Some skills here are shared with other agents I've published. They're bundled in each repo that needs them so every repo installs standalone. If you improve one of these, the same change may be worth applying wherever it appears:

• antigen-epitope-pipeline — also used by other agents in this collection

License

MIT — see LICENSE.