prototype-adr

Extract ADR prototype objectives and orchestrate controlled experiments across isolation backends.

1|Updated Apr 1, 2026
One-click install
npx skills add https://github.com/juliusl/adr-skills --skill prototype-adr
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: prototype-adr
Source: https://github.com/juliusl/adr-skills/tree/main/src/skills/prototype-adr
Command: npx skills add https://github.com/juliusl/adr-skills --skill prototype-adr

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) and assets (resource) components.

What problem does it solve?

Prototyping and validating architectural decisions by turning ADR objectives into controlled experiments and empirical evidence.

Core Features & Use Cases

  • Read ADRs produced by author-adr and extract prototype objectives from Evaluation Checkpoint and related sections.
  • Orchestrate experiments across isolated environments (worktree, container, or ACP sandbox) and record observations for auditability.
  • Feed findings back into the ADR lifecycle by updating context, strengths/weaknesses, and decision status.

Quick Start

Create a minimal profile and run a prototype by selecting an ADR and starting a controlled experiment.

Frequently Asked Questions about prototype-adr

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I validate architecture decisions with controlled experiments?

To validate architecture decisions, this Skill extracts prototype objectives from ADRs and orchestrates controlled experiments across isolated backends. It captures empirical evidence and feeds findings back into the ADR lifecycle by updating context and decision status.

How do I extract prototype objectives from an ADR evaluation checkpoint?

Extracting prototype objectives from an ADR involves reading the Evaluation Checkpoint and related sections. The Skill identifies these targets to guide experiments, translating architectural decisions into testable prototype goals.

Can I run ADR experiments in isolated environments like worktrees or containers?

You can run ADR experiments in isolated environments including worktrees, containers, and ACP sandboxes. The Skill orchestrates these isolation backends to execute controlled prototype experiments and record structured observations for auditability.

How do I record observations from ADR experiments for auditability?

Recording observations from ADR experiments is done using JSONL files and structured results. The Skill logs empirical data from controlled experiments in these formats to maintain auditability and feed findings back into the ADR.

Do I need TOML profiles to prototype architectural decisions across projects?

You need TOML profiles stored in the .adr/profiles directory to prototype architectural decisions. These profiles configure the orchestration of isolation backends and experiments across different projects.

What is the best way to feed experiment findings back into the ADR lifecycle?

The best way to feed experiment findings back into the ADR lifecycle is by using structured results to update context, strengths, weaknesses, and decision status. This ensures architectural decisions are continuously validated by empirical evidence.