Atlas

The Atlas blog

Building software with AI coding agents

What's new in AI coding, explained plainly. Guides for founders, product owners and testers who work with an AI team, and hands-on how-tos for the engineers who run Claude Code and OpenAI Codex.

RSS
  1. TechnicalFrom AtlasClaude Code Output Styles: Concise, Custom and CostClaude Code output styles change tone, length and role for a whole session. Learn the four built-ins, how to write a custom one, and what it does to cost. · 7 min read
  2. Plain EnglishExplainerDatabase Backups Before an AI Release: What to AskBefore an AI-built release changes your database, is the backup proven? A plain guide to backups, restore tests and rehearsals, with questions to ask. · 7 min read
  3. Plain EnglishWhat's newClaude Code Now Asks Before Sending Your Email OnlineClaude Code 2.1.317 (3 Nov 2026) asks before a web or shell tool call sends your account email to the network. What changed and what teams should do. · 5 min read
  4. TechnicalHow-toClaude Code Agent Teams: Setup, Cost and When to UseHow to enable Claude Code agent teams, choose models, gate work with hooks, control token cost, and what the experimental feature cannot do yet. · 9 min read
  5. Plain EnglishExplainerTechnical Debt in AI-Built Software: A Plain GuideWhat technical debt is, why AI-built software can pile it up quietly, and the plain steps a non-coder can take to keep it small and affordable. · 6 min read
  6. Plain EnglishWhat's newClaude Code Now Warns 10 Days Before Your Login ExpiresClaude Code 2.1.318 shows the date your claude.ai sign-in ends and warns 10 days ahead in scripted runs. What it means for overnight, unattended AI work. · 6 min read
  7. TechnicalHow-toClaude Code /compact vs /clear: Manage Context and CostWhen to use /clear, /compact or /rewind in Claude Code, what compaction keeps, and how the prompt cache changes what each command costs you in practice. · 9 min read
  8. Plain EnglishHow-toHow to Write a Spec With an AI Coding Agent (No Code)Not sure what you want built yet? Let the AI coding agent interview you, write a one-page spec, then check the finished work against it. A plain guide. · 8 min read
  9. Plain EnglishWhat's newClaude Code Now Fails Loudly When Automated Runs BreakClaude Code 2.1.314 and 2.1.317 fix automated claude -p runs that exited 0 after failing. What changed, why it matters to your team, and what to check. · 5 min read
  10. TechnicalHow-toClaude Code Resume: Continue, Name and Branch SessionsHow to resume a Claude Code session with --continue, --resume and /resume, name sessions, branch with /branch, and avoid a costly cold-cache resume. · 9 min read
  11. Plain EnglishExplainerDoes an AI Coding Agent Share Your Code? Plain GuideWhat happens to your code when an AI coding agent works on it: what leaves your computer, who may train on it, how long it is kept, and what you can switch off. · 7 min read
  12. Plain EnglishWhat's newCopilot Automations: Shared, Capped Background AgentsGitHub Copilot Automations now run as shared team workflows with spending caps and 24 new triggers. What changed on 28 October and what to check first. · 6 min read
  13. TechnicalHow-toClaude Code Model Config: Aliases, opusplan and EffortHow to choose the model in Claude Code: aliases, opusplan, effort levels, fallback chains, subagent models and the order in which settings win. · 8 min read
  14. Plain EnglishHow-toHow to Report a Bug to an AI Coding AgentFound a bug in AI-built software? Learn what to write, show and ask for so the agent fixes the right thing the first time, with no code skills needed. · 7 min read
  15. Plain EnglishWhat's newClaude Fast Mode Ends for Opus 4.8 on 30 NovemberFast mode for Claude Opus 4.8 is deprecated and ends 30 November 2026. See what changes, who is affected and what to do before then. · 6 min read
  16. TechnicalHow-toCodex config.toml: Profiles, Project Config, PrecedenceHow Codex config.toml layers work: user, profile, project and CLI settings, which one wins, what projects cannot set, and a safe setup for a team repo. · 5 min read
  17. Plain EnglishExplainerCan an AI Coding Agent Work Overnight? A Plain GuideCan an AI coding agent work while you sleep? Yes, if you answer five questions first. A plain-English guide to unattended runs, checks, limits and sign-off. · 7 min read
  18. Plain EnglishWhat's newClaude Code Ultrathink Now Raises Effort to xhighClaude Code's ultrathink keyword can now lift one turn to xhigh effort, and subagents get an effort setting. What changed and how to use it well. · 7 min read
  19. TechnicalHow-toCodex in CI: codex exec and the GitHub ActionRun Codex in CI with codex exec and openai/codex-action: read-only defaults, API key handling, safety strategies and a two-job pull request review workflow. · 6 min read
  20. Plain EnglishExplainerPrompt Injection Explained: A Plain Guide for FoundersPrompt injection is when text an AI agent reads tricks it into acting against you. Learn what it is, what could go wrong, and the habits that cut the risk. · 7 min read
  21. Plain EnglishWhat's newClaude Code Can Now Block Writes Outside Your ProjectClaude Code 2.1.314 (30 Oct 2026) adds a setting that makes its file-editing tools refuse paths outside your project in every mode. What it covers and misses. · 6 min read
  22. Plain EnglishWhat's newCopilot Fusion: Model Routing for AI Coding ExplainedGitHub Copilot Fusion went generally available on 28 Oct 2026. See how it picks AI models for you, what it costs, and what teams should check first. · 5 min read
  23. TechnicalFrom AtlasClaude Code Rules: Path-Scoped InstructionsSplit a bloated CLAUDE.md into .claude/rules files that load only for matching paths. Setup, glob patterns, limits, debugging and an audit step. · 7 min read
  24. Plain EnglishExplainerWho Owns AI-Written Code? A Plain Guide for FoundersWho owns code an AI agent writes for you? What US copyright guidance says, what your contract adds, and simple steps that protect your business. · 6 min read