Edison SinaniAI implementation architectBook a call

AI Team Playbook / Appendix B

Other tools, limits and sources

Everything under docs/ in the kit is tool neutral, and the instruction file is read by all four of these tools. What moves from tool to tool is where the settings, hooks, agents and automated review live.

September 16, 2026 · About 2 minutes

Contents

The playbook

  1. Overview

Before code

  1. 1Write the constitution first
  2. 2One instruction file, under 200 lines
  3. 3Prose is advice, hooks are enforcement
  4. 4Three gates before any code
  5. 5A walkable demo before any screen

Building

  1. 6One task, one worktree, one pull request
  2. 7Every contract has a file and two tests

Reviewing

  1. 8AI review first, and filtered
  2. 9Human review by risk tier
  3. 10Done means the evidence is in the pull request

Over time

  1. 11Measure five numbers from day one
  2. 12Protect the people who are still learning

Appendix

  1. AInstall checklist
  2. BOther tools, limits and sources
  3. CTightening the process
  4. DRun the agent with no entitlements

Get the kit

The same setup in other tools

Piece Claude Code Cursor GitHub Copilot Kiro Codex
Instruction file CLAUDE.md imports AGENTS.md AGENTS.md, plus .cursor/rules/*.mdc AGENTS.md, plus .github/copilot-instructions.md AGENTS.md, plus .kiro/steering/*.md AGENTS.md
Shared permissions .claude/settings.json Team rules in the dashboard Organization custom instructions .kiro/agents/*.json ~/.codex/config.toml
Hooks Before a tool, after a tool, on stop .cursor/hooks.json .github/hooks/*.json .kiro/hooks/*.json Not documented, September 2026
Agents .claude/agents/*.md .cursor/agents/*.md, also reads .claude/agents/ .github/agents/*.agent.md .kiro/agents/ .codex/agents/*.toml
Parallel work claude --worktree Cloud agents in isolated machines One agent per issue in Actions Waves from the task graph Worktrees by hand
Automated first pass ai-review.yml, or managed review with REVIEW.md Bugbot, .cursor/BUGBOT.md Copilot code review Use one of the others @codex review
Spec workflow /spec and /tasks Same skills as rules, or Spec Kit Spec Kit Built in, .kiro/specs/ Spec Kit

The docs/specs/ folder uses Kiro’s names on purpose. A team on Kiro can point Kiro at it or move it to .kiro/specs/

. GitHub Spec Kit works with all five tools [18], and its constitution, plan and

[P] markers are the same things as this kit’s constitution, design and task markers.

What the playbook does not solve

  • It does not make the team faster on its own. It makes the team measurable and stops the three failure modes the evidence shows most often.
  • The automated review action changes over time. Check its documentation before the first run and at each milestone.
  • The hooks assume a Unix shell and Python 3. A team on Windows runs them under WSL or rewrites them in PowerShell.
  • The numbers come from other companies’ setups. They justify the shape of the rules, not any prediction about your team.
  • The gates catch a badly specified feature. They do not catch the wrong feature.

All sources

  1. 1Faros AI, "The AI Engineering Report 2026: The Acceleration Whiplash", April 2026. Telemetry from 22,000 developers and more than 4,000 teams over two years. www.faros.ai/blog/ai-acceleration-whiplash-takeaways
  2. 2Faros AI, "The AI Productivity Paradox", July 2025. Telemetry from more than 10,000 developers across 1,255 teams. www.faros.ai/blog/ai-software-engineering
  3. 3Daniel Ogenrwot and John Businge, "AgenticFlict: A Large-Scale Dataset of Merge Conflicts in AI Coding Agent Pull Requests on GitHub", arXiv 2604.03551, April 2026. 107,026 pull requests. arxiv.org/abs/2604.03551
  4. 4Suzhen Zhong, Shayan Noei, Ying Zou and Bram Adams, "Human-AI Synergy in Agentic Code Review", arXiv 2603.15911, March 2026. 278,790 review conversations across 300 open source projects. arxiv.org/abs/2603.15911
  5. 5METR, "Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity", July 2025. Randomized controlled trial, 16 developers, 246 tasks. metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/
  6. 6Jai Lal Lulla et al., "On the Impact of AGENTS.md Files on the Efficiency of AI Coding Agents", arXiv 2601.20404, January 2026. 10 repositories, 124 pull requests. arxiv.org/abs/2601.20404
  7. 7Anthropic, "How AI assistance impacts the formation of coding skills", February 2026. Randomized controlled trial, 52 engineers. www.anthropic.com/research/AI-assistance-coding-skills
  8. 8Uber Engineering, "uReview: Scalable, Trustworthy GenAI for Code Review at Uber", 2025. www.uber.com/blog/ureview/
  9. 9HubSpot Engineering, "Automated Code Review: The 6-Month Evolution", 2026. product.hubspot.com/blog/automated-code-review-the-6-month-evolution
  10. 10Umut Cihan et al., "Automated Code Review In Practice", ICSE 2025, Software Engineering in Practice track. arXiv 2412.18531. arxiv.org/abs/2412.18531
  11. 11Zalando Engineering, "Agentic Engineering at Zalando: a snapshot", August 2026. engineering.zalando.com/posts/2026/08/agentic-engineering-at-zalando-a-snapshot.html
  12. 12Anthropic, "How Anthropic secures its AI-native software development lifecycle", July 2026. claude.com/blog/how-anthropic-secures-its-ai-native-software-development-lifecycle
  13. 13Anthropic, Claude Code documentation, "Best practices for Claude Code". code.claude.com/docs/en/best-practices
  14. 14Anthropic, Claude Code documentation, "Manage costs effectively". code.claude.com/docs/en/costs
  15. 15Google Cloud DORA, "State of AI-assisted Software Development", 2025. Survey of nearly 5,000 technology professionals. dora.dev/research/2025/dora-report/
  16. 16Farhan Thawar, VP of Engineering at Shopify, on "comprehension debt", as reported in Addy Osmani, "Comprehension Debt: The Hidden Cost of AI-Generated Code", O'Reilly Radar, 2026. www.oreilly.com/radar/comprehension-debt-the-hidden-cost-of-ai-generated-code/
  17. 17"Adoption and Impact of Command-Line AI Coding Agents: A Study of Microsoft's Early 2026 Rollout of Claude Code and GitHub Copilot CLI", arXiv 2607.01418, July 2026. arxiv.org/abs/2607.01418
  18. 18GitHub, Spec Kit release 1.0.0, 21 August 2026. github.com/github/spec-kit/releases
  19. 19AWS Kiro documentation, "Specs": requirements, design and tasks with approval gates. kiro.dev/docs/specs/
  20. 20Business Insider, report on JPMorgan's Claude Code spending limits and its Devspace sandbox, September 2026. A 2,000 dollar monthly cap per engineer, a containerized environment on AWS, and CISO Pat Opet on agents with no entitlements. Paywalled; the figures here are as reported by the outlets that covered it. www.businessinsider.com/jpmorgan-claude-spending-limit-security-engineers-2026-9
  21. 21Anthropic, Claude Code documentation, "Configure the sandboxed Bash tool". code.claude.com/docs/en/sandboxing
  22. 22Anthropic, Claude Code documentation, "Development containers": the reference container, its egress firewall, and managed settings. code.claude.com/docs/en/devcontainer
  23. 23Anthropic, Claude Code documentation, "Monitoring": OpenTelemetry export of usage, cost and tool activity. code.claude.com/docs/en/monitoring-usage