Shootquinn

CR Agents — Testing skill for Claude Code

Testing community

Brute-force AI method for technical work.

How to install CR Agents

This entry records only its repository, not the path inside it, so there is no exact command to give. Open Shootquinn/CR-Agents and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What CR Agents does

Brute-force AI method for technical work. Personas anchored to experts in their fields run as Claude Code sub-agents with clean context windows, test-driven quality gates, and accumulated judgment (pseudo-ego) across sessions.

Alternatives in Testing

Deslop Reborn — One iteration of the IronClaw Reborn de-slop loop — take ONE Reborn crate, fan out parallel review sub-agents 12.6k ★
AI Contribution Skills (Engram) — branch-pr: clean branch and PR workflow 2.5k ★
Cc10x — The Loop Engine for Claude Code — engineer the loop, not the prompt 163 ★

README

Collaborative Reasoning + Test-Driven Documentation

By: Quinn Morley, Unleashed Robotics

A method for performing technical work using AI-simulated expert teams, test-driven quality assurance, and (optionally) programmatic CAD. Runs on Claude Code. Applicable to engineering documents, proposals, hardware design, systems analysis, or anything complex enough that getting it wrong is expensive.

Based on Nelson (2026), "Activating Expertise: Human-AI Teams for Collaborative Reasoning," implemented via spawned sub-agents with clean context windows per persona, with accumulated judgement calls to enable a persistent "pseudo-ego." Nelson's method is further augmented with the Test Driven Documentation method developed by the author in late 2025.

Quickstart

Point Claude Code at CLAUDE.md. The entry point for every session is:

Read CLAUDE.md. Work through prompt0. Here is my work order: [paste work order or point to file]

Or more simply:

Read CLAUDE.md. Work through prompt0. Today we're going to design an airplane.

That is the entire entry point. CLAUDE.md bootstraps the session: it tells the orchestrator where to find the operational guide, the gameplan, and the accumulator. prompt0 verifies tools and establishes session capabilities. The team assembles, builds a gameplan for your review, writes a test suite, and begins executing. You approve each step before the next one starts.

After compaction or a new session, the recovery incantation is the same: "Read CLAUDE.md, work through prompt0, and read the gameplan."

LLM-PLM (Product Lifecycle Management performed by the LLM) activates when CAD/geometry creation is expected to occur. If CAD work happens later, the team will pick up the supplement automatically. You can read more about this in the `supplements/llm_plm_cad.md` file.