cc-safe-setup
Listed on Product Hunt since April 21, 2026.
One command to make Claude Code safe for autonomous operation. 897 example hooks · 73+ Anthropic Issues addressed by hook · 218 test files · 30K+ npm downloads (cumulative) · 日本語
npx cc-safe-setup
Installs 8 safety hooks in ~10 seconds. Blocks rm -rf /, prevents pushes to main, catches secret leaks, validates syntax after every edit. Zero npm dependencies. Hooks use jq at runtime (brew install jq / apt install jq).
What's a hook? A checkpoint that runs before Claude executes a command. Like airport security, it inspects what's about to happen and blocks anything dangerous before it reaches the gate.
What problem are you solving today? — routing informed by the 6 readers who actually bought the books on the right column
| Your situation | Start here (free) | Go deeper (¥800-$19) |
|---|
Stop destructive ops (rm -rf, force-push, prod commands) | Live Demo → npx cc-safe-setup | 事故防止本 (¥800) — the 6-buyer entry book |
Claude merged or deployed to production without your approval — gh pr merge --admin bypassed branch protection, or a deploy fired autonomously (#68676) | gh-cli-destructive-guard + deploy-guard — PreToolUse on Bash, refuse gh pr merge / deploy commands at the tool boundary even when --admin overrides server-side branch protection · npx cc-safe-setup | 事故防止本 — JP (¥800) — the irreversible-ops prevention book |
Uncommitted work silently gone — no command you ran, no error (a git reset --hard you didn't type, harness checkpointing, crash auto-stash) | Recover it — your edits are usually still on the stash stack: git stash list → git stash apply · Incident Tracker — Data Loss · npx cc-safe-setup's uncommitted-discard-guard blocks the model-driven variant | 事故防止本 — JP (¥800) — the data-loss prevention book · Autonomous Claude Ops — EN (~$6) — Ch.7 is this exact failure |
Token costs out of control (/cost shock) | Token Checkup | Token Book — JP (¥2,500) · Token Book — EN (name your price) — the staged-upgrade destination |
| Max plan hitting usage limits too fast (rate-limited at low %) | Why Max hits limits so fast (the #16157 cluster + what you can measure) · Token Drain Analyzer (analyze your own logs in-browser → bug or context?) | Token Book — JP (¥2,500) · Token Book — EN (name your price) — cut consumption |
| June 15 billing cliff — Pool 2 splits programmatic usage into a separate paid credit pool, so the same automation can cost 5–25× more (two model IDs also hard-404 the same day) | Exposure from your logs — whether it hits you, read from your real session logs (not a guess) · Credit Cliff Calculator — how much you'll overage · 4-Path Picker — which response path fits you | 6月15日の課金分離に備える (¥800) — 4 operator paths, day-by-day timeline, the M metric |
