Claude Code mod

guardrails

Safety and review of Claude's output, helped by local decision models. Mods: blast-radius, secret-sentry, reflect (/reflect-queue, /reflect-pause), assumption-tracker (/assumptions), self-critique (/critique), prompt-coach (/sharpen), merge-sweep, decision-gate (/decide-bench), auto-answer, model-picker. Skill: guardrails-setup (/guardrails:guardrails-setup).

Guard & safety mods Plugin marketplaceValidates on Claude Code 2.1.322 Source

Install

Inside Claude Code
/plugin marketplace add bharath-gp/ai-harness
/plugin install guardrails@ai-harness

Mods install like plugins. Run /reload-plugins or start a new session to load it.

What it can access

Reach
Unknown
Sees
Bash calls, Write calls, Edit calls, NotebookEdit calls, classic hooks, AskUserQuestion calls, subagent spawns, every prompt
Draws in
AskUserQuestionAbovePromptPane
Hooks
ui.faultui.rendertool.calltool.checkclassic.UserPromptSubmitagent.spawncommand.runsession.startprompt.submitturn.complete

From a static scan of the mod's code by awesome-claude-code-mods. It shows what the code asks for, not a security review.

How Claude Code mods work

  • A mod is a plugin with a JavaScript or TypeScript module that runs inside Claude Code (version 2.1.287 or later). It can draw panes beside the transcript, a band above the prompt, buttons and toasts, restyle the spinner and tool rows, and watch or rewrite tool calls and prompts.
  • It installs like any plugin: add its marketplace with /plugin marketplace add owner/repo, then /plugin install name@marketplace.
  • Check what a mod can reach before installing: it runs with your permissions. Reach goes from drawing only, to reading files, to writing files or running processes, to the network. Run claude plugin validate on a clone to list its hooks and calls without running it.
  • Turn one off from the Installed tab in /plugin, or start Claude Code with --safe-mode to skip every customization for a session.
Claude Code docs

Details

Repository
bharath-gp/ai-harness
Folder
plugins/guardrails
Marketplace
ai-harness
Type
Guard & safety mods
License
MIT
Last updated
6 November 2026
Listed in
awesome-claude-code-mods

FAQ

What is guardrails?

Safety and review of Claude's output, helped by local decision models. Mods: blast-radius, secret-sentry, reflect (/reflect-queue, /reflect-pause), assumption-tracker (/assumptions), self-critique (/critique), prompt-coach (/sharpen), merge-sweep, decision-gate (/decide-bench), auto-answer, model-picker. Skill: guardrails-setup (/guardrails:guardrails-setup). It is a mod for Claude Code, from bharath-gp/ai-harness on GitHub.

What can the guardrails mod access?

Its reach could not be worked out from its code. It sees Bash calls, Write calls, Edit calls, NotebookEdit calls, classic hooks, AskUserQuestion calls, subagent spawns, every prompt. It draws in AskUserQuestion, AbovePrompt, Pane. Like any mod, it runs with your permissions.

How do I install guardrails?

Inside Claude Code (v2.1.287 or later), run /plugin marketplace add bharath-gp/ai-harness and then /plugin install guardrails@ai-harness.

More guard & safety mods

guardian

Juan-Gpu/mis-mods

Pide confirmación con botones antes de que Claude ejecute comandos peligrosos (rm -rf, git push --force, git reset --hard)

Guard & safety mods/plugin install

guardrails

juniormartinxo/jm-claude-mods

Travas dos fluxos de card: entrega sem commit/push/PR, comentários fora do pt-BR, Linear em projeto do It's a Plan e card criado sem o consentimento do usuário

Guard & safety mods

guardia-produccion

Vincitorecode/mods-equipo

Detiene comandos riesgosos (deploys, migraciones, borrados, force push, kubectl en prod) y pide confirmación; /guardia muestra lo que vigila y lo que detuvo. · Stops risky commands (deploys, migrations, deletes, force push, kubectl on prod) and asks first.

Guard & safety mods/plugin install

guardrails

Stoica-Mihai/claude-mods

Blocks the Bash commands your own rules file names, and shows every block in /blocked. Ships no rules.

Guard & safety mods/plugin install

guardia-llaves

enpixeles-ai/mods-claude

Tapa las llaves y contrasenas antes de que Claude las vea, y le prohibe imprimir archivos de secretos.

Guard & safety mods/plugin install

guardrails-approval

scaso01/claude-risk-guardrails

A human signs off before anything that can't be undone: publishing, killing protected processes, editing secrets or stale copies.

Guard & safety mods/plugin install

Browse the most popular Claude Code mods, all guard & safety mods, or the skills and MCP server directories.