Agent Runway — Testing skill for Claude Code
Operational discipline layer for AI-assisted software delivery: spec/ticket workflows, quality gates, and persistent engineering memory across Cursor, Claude Code, and VS Code.
How to install Agent Runway
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open adonai-labs/agent-runway and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Agent Runway does
Operational discipline layer for AI-assisted software delivery: spec/ticket workflows, quality gates, and persistent engineering memory across Cursor, Claude Code, and VS Code.
Alternatives in Testing
- AI Contribution Skills (Engram) — branch-pr: clean branch and PR workflow 2.5k ★
- Test Driven Development.SKILL — Enforces TDD discipline with RED-GREEN-REFACTOR cycle 262 ★
- GAAI Framework — 95+ Drop-in governance layer (.gaai/ folder) -- backlog authorization, cross-session memory, decision tracking 113 ★
README
Agent Runway
[npm](https://www.npmjs.com/package/@adonai-labs/agent-runway) · [license](https://github.com/adonai-labs/agent-runway/blob/HEAD/LICENSE) · [repository](https://github.com/adonai-labs/agent-runway) · [roadmap](https://github.com/adonai-labs/agent-runway/blob/HEAD/ROADMAP.md)
**AI agents are good at writing code. They are not good at remembering what they decided last week.**
Agent Runway keeps delivery knowledge in the repository: specs, memory, checkpoints, run logs, rules, and quality gates. Agents and humans can keep building without starting from scratch every session.
It does not replace your stack, architecture, or process. It wraps around them.
As of v1.6, Agent Runway supports Cursor, Claude Code, and VS Code Copilot, and ships implementation skills for .NET, TypeScript, Node.js, React, Python, Go, and Rust.
Requirements
- Node.js 22 or newer
Quick Start
Choose the amount of structure you want.
Vibe/Lite
Fast, skill-first setup for prototypes, solo work, and low-ceremony coding.
npx @adonai-labs/agent-runway init --preset vibe-lite --target all
Lite installs `@start`, `@express`, `@code-review`, `@checkpoint`, `@learning`, `@safety-check`, `@council` preview, and a small rule set for engineering principles, security, and testing. It does not install slash command aliases, agents, or heavy structured delivery workflows.
When the work outgrows Lite, upgrade in place:
npx agent-runway upgrade --to structured
This preserves memory, logs, checkpoints, and docs while adding the full Structured workflow set.
Structured
Full delivery workflow for long-lived products, teams, and work that needs traceability.
npx @adonai-labs/agent-runway init --preset node-typescript --target all
Structured installs slash commands, agents, specs, tickets, lead/autonomous workflows, architecture/contrarian gates, `@council` preview, full rules, and optional governance checks.
Or install first:
Related Skills
AI Native Playbook
The harness I use to run coding agents on regulated production codebases: agent contracts, skills as runbooks,
Orbit Mode
A multi-agent AI engineering control center for orchestrating Claude, Codex, and other coding agents across Gi
Hoofy
Hoofy — AI development companion MCP server. Persistent memory, spec-driven development, adaptive change pipel
AI Assisted Testing Demo
Companion repo for "Just Enough Context: Teaching Claude to test with Playwright". Three context layers (AGENT
Cc Codex Triage
Claude Code plugin for persistent triage dialogue with the OpenAI Codex CLI via codex exec resume — cross-agen
Tatai
Tatai 塔台 — local-first workbench for one person orchestrating multiple AI coding agents (Claude Code / Codex /
Related Agents
Wick
Wick — persistent thinking partner with philosophical frameworks, operational gates, calibration discipline, a
Delivery Tpm
Delivery control (code: tpm). Reconciles the board with branches, worktrees, pull requests, and gates; keeps o
Reliability Engineer
Python delivery agent for quality gates and test infrastructure. Owns test suite (>=80% coverage enforced), go