Skip to content

README.md

Latest commit

 

History

History
99 lines (76 loc) · 3.79 KB

README.md

File metadata and controls

99 lines (76 loc) · 3.79 KB

Refua and ClawCures Guidebook

This guidebook is a textbook-style manual for the AgentCures ecosystem. It teaches both the science and the system design, so readers can move from "I have a drug discovery objective" to "I have an executable, auditable campaign."

Who This Is For

  • Research engineers learning discovery workflows
  • Computational scientists building model pipelines
  • Platform engineers operating campaign infrastructure
  • Wet-lab and preclinical teams integrating operational evidence
  • Clinical strategy teams planning trial simulations
  • QA, governance, and regulatory stakeholders reviewing readiness

How To Use This Guidebook

Pick a reading track:

  • Fast onboarding track: Chapters 1, 2, 4, 8
  • Science depth track: Chapters 1, 3, 5, 9, 10
  • Operations track: Chapters 2, 5, 7, 8
  • Governance track: Chapters 5, 6, 8, Appendix B

Narrative Arc

If you read straight through, the guidebook follows one continuous story:

  1. understand how discovery decisions are actually made under uncertainty
  2. learn how the platform turns intent into typed, executable workflows
  3. run an end-to-end campaign with evidence and gate controls
  4. deepen into medicinal chemistry and development-stage decision science

Chapter Index

  1. Chapter 1: Drug Discovery For Builders
  2. Chapter 2: Platform Architecture
  3. Chapter 3: Refua Core Science Engine
  4. Chapter 4: ClawCures Campaign Orchestrator
  5. Chapter 5: Program Lifecycle Modules
  6. Chapter 6: Quality, Governance, and Evidence
  7. Chapter 7: Deployment and Runtime Operations
  8. Chapter 8: End-to-End Walkthrough
  9. Chapter 9: Medicinal Chemistry and Molecular Optimization
  10. Chapter 10: Drug Discovery and Development Science
  11. Appendix A: Glossary
  12. Appendix B: Artifacts and Schemas

Data And Rich Artifacts

Use these artifacts while reading:

Ecosystem In One Diagram

flowchart TD
    U[Team Objective] --> C[ClawCures Planning]
    C --> M[refua-mcp Typed Tools]
    M --> R[refua Modeling]
    M --> D[refua-data]
    M --> W[refua-wetlab]
    M --> P[refua-preclinical]
    M --> L[refua-clinical]
    C --> H[refua-schema Portfolio Model]
    C --> S[clawcures-ui]
    S --> B[refua-bench]
    S --> G[refua-regulatory]
    X[refua-deploy] --> C
    X --> M
    X --> S
Loading

Teaching Style Used In Chapters

Each chapter follows a consistent layout:

  • learning goals
  • conceptual model
  • concrete package mapping
  • practical checklists
  • common failure modes
  • chapter handoff to next topics

Conventions

  • "Campaign" means one executable objective with traceable artifacts.
  • "Tool call" means a typed JSON request handled through refua-mcp.
  • "Gate" means a formal quality or policy checkpoint.
  • "Evidence" means reproducible and integrity-verifiable outputs.

Important Scope Note

This guidebook describes software, simulation, and operational workflows. It does not claim therapeutic efficacy or regulatory approval outcomes.