A PreToolUse hook is the guardrail that survives bypass mode
Permission prompts depend on what mode you're in. A PreToolUse deny hook doesn't. Here's a small one that blocks force pushes, and where it falls short.
3 min read
Assistants
Coding assistants and agents, and how to get better results from them.
A separate evaluator agent is what makes long agent runs worth the money
Anthropic's harness splits planning, building and testing across agents. The cost jump was 20x, and the reasons it paid off are specific.
3 min read
Cursor Projects puts a coordinator agent in charge of the work
The agent that plans and delegates never writes the code. Cursor says it can keep a project's context for months.
3 min read
What 16 parallel agents building a C compiler teach about running agents at scale
Anthropic's engineering team spent nearly $20,000 on it. The lessons about tests, locks and duplicate code transfer to much smaller jobs.
3 min read
Let the agent interview you, then write the spec, then start fresh
A three-step workflow for features too big for one prompt, and what makes a spec worth handing over.
3 min read
Claude Code's auto mode catches most risky actions. Anthropic's numbers show what it misses
A 0.4% false positive rate sounds great. The 17% miss rate on real overeager actions is the number to plan around.
3 min read
Run parallel Claude Code sessions without them stepping on each other
The --worktree flag gives each session its own checkout and branch. The setup details are where people get caught.
3 min read
Your CLAUDE.md is probably too long. Here's how to cut it
Anthropic's own advice is to prune ruthlessly. A one-question test, a before and after, and where each deleted rule should go instead.
3 min read
How Codex reads AGENTS.md, and how to layer it so the right rules win
Global defaults, repo rules and per-service overrides stack in a fixed order. Know it, and a monorepo stops fighting your instructions.
3 min read
Stop telling your AI agent to 'fix the bug'. Make it reproduce it first
A bug prompt in three parts, and a rule for when to throw the whole session away.
4 min read
Make Claude Code run your tests before it's allowed to say it's done
A Stop hook turns 'please run the tests' from a request into a rule. Here's a working setup and the trap to avoid.
3 min read
Cursor's Rollouts bot watches your deploys after the merge
Most AI tooling stops at the pull request. This one keeps going until the change is live in each environment.
3 min read