AtlasThe 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.
- TechnicalHow-toClaude Code Permissions: Allow, Deny Rules and ModesSet up Claude Code permissions that hold: how allow, ask and deny rules are ordered, which mode to pick, and the Bash rule traps that let commands slip through.
- Plain EnglishHow-toHow to Brief an AI Coding Agent: A Plain-English GuideHow to brief an AI coding agent without writing code: what to put in the request, how to define "done", and the common mistakes that cost you rework.
- Plain EnglishWhat's newClaude Code Idle Compaction: Cheaper Breaks, Same SessionClaude Code 2.1.299 (14 Oct 2026) now compacts very long idle chats before the one-hour cache runs out. What changed, the trade-off, and how to turn it off.
- TechnicalExplainerClaude Code Skills: Turn Repeated Prompts into CommandsClaude Code skills turn a checklist you keep pasting into a reusable /command. Covers SKILL.md, arguments, shell injection and tool permissions.
- Plain EnglishHow-toHow to Check AI-Built Software When You Don't Write CodeYou can't read the code, but you can still tell whether an AI agent's work is good. Ask for evidence, set pass/fail checks, and try the edge cases yourself.
- Plain EnglishWhat's newClaude Haiku 5.5: What the Cheapest Claude Means for TeamsClaude Haiku 5.5 arrived on 7 October 2026 with a 1M-token window at $0.10 per million input tokens. What it is good for, what it is not, and how to test it.
- TechnicalExplainerMCP Servers in Claude Code: Connect Databases and ToolsHow MCP servers work in Claude Code: add local and remote servers, choose a scope, handle OAuth and secrets, limit context use and lock them down for a team.
- TechnicalHow-toClaude Code Costs: Cut Token Usage and Stretch LimitsWhere Claude Code spends tokens and how to cut it: /usage, model and effort choice, /clear vs /compact, prompt cache lifetime and team spend limits.
- TechnicalHow-toAI Code Review with Claude Code: What It Catches and MissesSet up AI code review in Claude Code with /code-review, ultrareview, managed Code Review or GitHub Actions, and learn what it catches, misses and costs.
- TechnicalHow-toClaude Code Headless in CI: -p Mode and Safe PermissionsRun Claude Code headless in CI with claude -p: bare mode, dontAsk and allowedTools permissions, JSON output, exit codes and cost caps for unattended jobs.
- TechnicalHow-toClaude Code Subagents: When to Split Work and HowWhen Claude Code subagents help and when they don't: writing the agent file, choosing tools and models, and catching frontmatter mistakes that fail silently.
- TechnicalHow-toClaude Code Hooks as Guardrails: Block, Format, TestUse Claude Code hooks to block risky shell commands, protect files, format every edit and keep Claude working until tests pass, with configs you can copy.
- TechnicalComparisonClaude Code vs Codex for Teams: A Workflow ComparisonClaude Code vs OpenAI Codex for engineering teams: instruction files, permissions, hooks, CI, cloud tasks and PR review compared, and when to pick which tool.
- TechnicalHow-toCodex CLI Setup: Install, Approval Modes and SandboxingSet up OpenAI Codex CLI: install it, sign in, write a config.toml and profile files, and choose the sandbox mode and approval policy for each kind of task.
- TechnicalHow-toCLAUDE.md Best Practices: What to Put in Your Project FileWhat belongs in CLAUDE.md, what to cut, and when to use rules, imports, skills and hooks instead, based on Anthropic's Claude Code documentation.
- TechnicalHow-toClaude Code Worktrees: Run Parallel Sessions SafelyHow to run parallel Claude Code sessions with git worktrees: the --worktree flag, .worktreeinclude, subagent isolation, cleanup rules and merging back.
Nothing here yet — new articles arrive three times a day.