> Security hooks (0 tokens) + dev pipeline + smart defaults. Install in one command, save 85% token overhead_
> Bash scripts outside context. Zero token cost. They enforce, not guide._
Blocks rm -rf /, force push to main, DROP TABLE, npm publish.
Detects role override attempts and manipulation patterns in real-time.
Warns when API keys (Anthropic, OpenAI, AWS, GitHub, Stripe) appear in output.
Blocks read/write to .env, credentials.json, SSH keys, .npmrc.
Saves git state and active plans before context compaction. Recoverable context.
If behavior can be enforced with a hook (0 tokens), don't write a rule. Hooks enforce. Rules guide.
> On-demand tools. Zero tokens when idle. Load only when invoked._
/initAuto-generate project CLAUDE.md
/discoverSurface unknowns before you plan
/specPlan a change into an approvable artifact
/craftBuild the approved plan, review, ship
/vetClean-context review: SHIP/HOLD/BLOCK
/exerciseBehavioral gate: drive the app as a user
/git-saveConventional commits + safety
/slop-scanDetect AI slop and tech debt
/htmlPlanViewer / ResultViewer artifacts
> Get started in 3 commands_
$ curl -fsSL https://raw.githubusercontent.com/brolag/neural-claude-code/main/install.sh | bash
> /init
Installs 5 hooks, 9 skills, 5 rules. Merges into existing settings.json without overwriting.
> /plugin marketplace add brolag/neural-claude-code
> /plugin install neural@neural-claude-code
Skills namespaced as /neural:spec, /neural:craft, etc. Pick ONE install method — the plugin and the curl installer together double-fire the hooks. The plugin ships hooks + skills; the curl installer also adds the compact rules and CLAUDE.md template.
Hooks + skills + rules. ~635 tokens/msg. Zero dependencies. Production-grade guardrails in one command.
CLONE_REPO