Deterministic, artifact-first protein design runtime and CLI
Project description
Agentic Proteins
Deterministic, artifact-first protein design runtime and CLI for traceable runs, human review, and reproducible decision trails.
At a glance: Deterministic runs • artifact-first output • human-in-the-loop gating • stable JSON contracts
Audience: Computational biology teams and lab workflows that require auditability and reproducibility.
Table of Contents
- Why Agentic Proteins?
- Try It in 30 Seconds
- Installation
- Quick Start
- What This Does
- What This Does Not Do
- Stable Public Surface
- Docs & Resources
- Contributing
- License
Why Agentic Proteins?
Protein design workflows need traceable decisions, stable artifacts, and reproducible execution paths. Agentic Proteins provides a deterministic loop that separates planning, execution, evaluation, and human decision points so that labs can audit and reproduce runs without digging into internal code.
Try It in 30 Seconds
pipx install agentic-proteins # Or: pip install agentic-proteins
agentic-proteins run --sequence "ACDEFGHIKLMNPQRSTVWY"
agentic-proteins inspect-candidate --latest
Installation
Requires Python 3.11+.
# Isolated install (recommended)
pipx install agentic-proteins
# Standard
pip install agentic-proteins
Quick Start
# Run a deterministic loop
agentic-proteins run --sequence "ACDEFGHIKLMNPQRSTVWY"
# Resume a run by ID
agentic-proteins resume --run-id <run_id>
# Compare two runs
agentic-proteins compare --left <run_id> --right <run_id>
What This Does
- Runs a deterministic agentic loop with traceable artifacts and telemetry.
- Produces proxy structure-quality signals from sequence heuristics by default.
- Requires explicit opt-in to run real structure predictors via
--provider.
What This Does Not Do
- It does not run real structure prediction unless
--provider esmfold|rosettafold|openproteinis specified and requirements are met. - It does not finalize candidate selection without a signed human decision artifact.
- It does not expose internal modules as stable APIs.
Stable Public Surface
- Supported CLI commands:
run,resume,compare,inspect-candidate,export-report. - Stable artifacts:
artifacts/<run_id>/layout and JSON payloads. - Stable CLI JSON output schema (see docs).
Docs & Resources
Start here:
-
Getting Started: https://bijux.github.io/agentic-proteins/overview/
-
CLI Reference: https://bijux.github.io/agentic-proteins/cli/cli/
-
Architecture: https://bijux.github.io/agentic-proteins/architecture/architecture/
-
Concepts: https://bijux.github.io/agentic-proteins/concepts/terminology/
-
Changelog: https://github.com/bijux/agentic-proteins/blob/main/CHANGELOG.md
-
Repository: https://github.com/bijux/agentic-proteins
-
Security: https://github.com/bijux/agentic-proteins/security/advisories/new
Contributing
See CONTRIBUTING.md for setup, style, and tests.
License
Apache-2.0 — see LICENSE. © 2025 Bijan Mousavi.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file agentic_proteins-0.1.0.tar.gz.
File metadata
- Download URL: agentic_proteins-0.1.0.tar.gz
- Upload date:
- Size: 5.5 MB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.1.0 CPython/3.13.7
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
5fa21bb135ba39d065a27f2f82363c33b60c62dbfdb0e912ec81e23ce1efdea3
|
|
| MD5 |
3fde6db171ddcb96f696cddfe87f9c68
|
|
| BLAKE2b-256 |
e68c2f18af7e3a4bd55ee379a6b4187c298c757dba84b0b87a571ca11020ef2d
|
File details
Details for the file agentic_proteins-0.1.0-py3-none-any.whl.
File metadata
- Download URL: agentic_proteins-0.1.0-py3-none-any.whl
- Upload date:
- Size: 143.7 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.1.0 CPython/3.13.7
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
cdee573ca46408c6ec37047c26b23864057a99ab7f06c503aa390fb99e3a012a
|
|
| MD5 |
9e25b45144cd8ae501db5a88b643ad9e
|
|
| BLAKE2b-256 |
4fa483b2260c06f221d39bb2cc4caa6e835caa0150fdbddf7759ec0eb8a413b8
|