git-guardrails-claude-code

by mattpocock

git-guardrails-claude-code adds a PreToolUse hook to block dangerous git commands before Claude Code runs them. Install it to prevent destructive pushes, hard resets, forced cleans, and branch deletes, with scoped control for this project or all projects. Useful when you need git-guardrails-claude-code for Access Control boundaries in Claude Code.

Stars66k

Favorites0

Comments0

AddedMay 8, 2026

CategoryAccess Control

Install Command

npx skills add mattpocock/skills --skill git-guardrails-claude-code

Curation Score

This skill scores 78/100, which means it is a solid listing candidate for Agent Skills Finder: directory users get a concrete safety workflow for blocking destructive git commands in Claude Code, with enough implementation detail to install and use it without starting from scratch. It is useful, but users should still confirm their scope choice and environment fit before adopting it.

78/100

Strengths

Explicit trigger and purpose: block dangerous git commands in Claude Code, including push, reset --hard, clean -f, and branch -D.
Operationally actionable: includes a step-by-step setup flow plus a bundled hook script at scripts/block-dangerous-git.sh.
Good agent leverage: uses a PreToolUse hook and shows exact settings.json snippets for project or global installation.

Cautions

Some setup assumes shell/tooling familiarity: users must copy the script, run chmod +x, and edit Claude settings correctly.
The doc excerpt is strong on blocking behavior but light on edge-case guidance and verification steps, so adoption may require some trial-and-error.

Claude Code Claude Git Bash Automation Hooks Rules

Overview

Overview of git-guardrails-claude-code skill

What git-guardrails-claude-code does

The git-guardrails-claude-code skill adds Claude Code hooks that block dangerous git commands before they run. It is aimed at teams or solo users who want an enforcement layer, not just a polite warning, for destructive actions like git push --force, git reset --hard, git clean -f, and git branch -D.

Who should install it

Use this git-guardrails-claude-code skill if you work in repositories where accidental data loss, accidental pushes, or local history rewrites are a real risk. It is especially useful when you want Claude Code to operate inside safer boundaries for Access Control and you need those boundaries enforced consistently.

What makes it different

Unlike a prompt-only safety reminder, this skill installs a PreToolUse hook that stops the command before execution. That means the guardrail is operational, not advisory. The main tradeoff is that it only protects the commands it matches, so it is best for clear, known-bad git actions rather than general policy enforcement.

How to Use git-guardrails-claude-code skill

Install the skill and choose scope

Install with:
npx skills add mattpocock/skills --skill git-guardrails-claude-code

The first decision is scope: install for this project only or for all projects. Project scope is safer when you want to test the hook in one repo first; global scope is better if your workflow should always block these commands across Claude Code.

Read the right files first

Start with SKILL.md to understand the setup flow, then inspect scripts/block-dangerous-git.sh because that is the enforcement logic. If you are adapting the behavior, the script matters more than the prose. This is the fastest path for a practical git-guardrails-claude-code guide.

How to form a good request

The skill works best when you specify where it should apply and what you consider dangerous. A strong request looks like: “Install git-guardrails-claude-code for this project only and block push, reset --hard, clean -fd, and branch -D.” A weaker request like “make git safer” leaves the scope and policy too vague to install cleanly.

Verify and customize the hook

After copying the hook to .claude/hooks/block-dangerous-git.sh or ~/.claude/hooks/block-dangerous-git.sh, make it executable and register it in the matching settings.json. Then test a blocked command in a safe repo to confirm Claude receives the refusal message. If your team uses additional risky commands, extend the pattern list in the script rather than relying on prompt text.

git-guardrails-claude-code skill FAQ

Is this a replacement for git permissions?

No. git-guardrails-claude-code is a local Claude Code guardrail, not a server-side access control system. It can stop Claude from executing matched commands, but it does not replace branch protections, repo permissions, review rules, or remote policy.

When is this a bad fit?

Skip it if you need unrestricted git automation, or if your workflow depends on scripts that legitimately use commands like git clean -fd or git branch -D. In those cases, a broad block may create more friction than value, and a narrower rule set may be better.

Is it beginner-friendly?

Yes, if you can follow a small install sequence and edit JSON. The concept is simple, but the best results come from understanding exactly which commands you want blocked and where the hook should live. Beginners usually struggle most with scope selection, not with the hook itself.

Does ordinary prompt engineering do the same thing?

No. Prompting can ask Claude to behave cautiously, but it cannot reliably enforce command blocking. The git-guardrails-claude-code skill is useful when the decision to stop a command must happen even if the model would otherwise comply.

How to Improve git-guardrails-claude-code skill

Tighten the blocked command list

The biggest quality gain comes from matching your actual risk profile. Review whether your team needs only git push and git reset --hard, or whether git clean -f, git checkout ., and git restore . should also be blocked. Better inputs produce fewer false positives and fewer gaps.

Keep the scope deliberate

Use project scope for experiments, shared-repo onboarding, or one-off protection in a sensitive codebase. Use global scope only when you are confident the same policy should follow you everywhere. This is the main decision that changes adoption success for the git-guardrails-claude-code install.

Test the failure mode

After installation, try a command that should be blocked and confirm the refusal is clear and immediate. If the hook is too permissive, harden the pattern list; if it blocks too much, narrow the patterns or separate project-specific rules from personal rules. Treat the first run as validation, not final state.

Ratings & Reviews

No ratings yet

Share your review

0/10000

Latest reviews

Saving...

more skill

azure-identity-py

by microsoft

azure-identity-py helps set up Azure authentication in Python with Microsoft Entra ID. Use it to choose DefaultAzureCredential, managed identity, or service principal auth, configure environment variables, and troubleshoot access control and credential chain issues. Install guidance, usage patterns, and practical setup notes are based on the repo skill file.

Access Control

Favorites 0GitHub 2.2k

django-security

by affaan-m

django-security is a practical guide for hardening Django apps with authentication, authorization, CSRF, XSS, SQL injection prevention, secure cookies, and production settings. It helps developers and reviewers run a focused Security Audit, quickly spot risky config, and apply concrete fixes before deployment.

Security Audit

Favorites 0GitHub 156.1k

detecting-anomalous-authentication-patterns

by mukul975

detecting-anomalous-authentication-patterns helps analyze authentication logs for impossible travel, brute force, password spraying, credential stuffing, and compromised-account activity. Built for Security Audit, SOC, IAM, and incident response workflows with baseline-aware detection and evidence-backed sign-in analysis.

Security Audit

Favorites 0GitHub 0

auditing-kubernetes-cluster-rbac

by mukul975

auditing-kubernetes-cluster-rbac helps audit Kubernetes RBAC for overbroad roles, risky bindings, secret access, and privilege escalation paths. It is built for security audit workflows across EKS, GKE, AKS, and self-managed clusters, with practical guidance for kubectl, rbac-tool, KubiScan, and Kubeaudit.

Security Audit

Favorites 0GitHub 0

auditing-gcp-iam-permissions

by mukul975

auditing-gcp-iam-permissions helps review Google Cloud IAM access for risky bindings, primitive roles, public access, service account exposure, and cross-project paths. This access-control audit skill is built for evidence-driven reviews with gcloud, Cloud Asset, IAM Recommender, and Policy Analyzer.

Access Control

Favorites 0GitHub 0

auditing-aws-s3-bucket-permissions

by mukul975

The auditing-aws-s3-bucket-permissions skill helps you audit AWS S3 buckets for public exposure, overly permissive ACLs, weak bucket policies, and missing encryption. Built for Security Audit workflows, it supports a repeatable least-privilege review with AWS CLI and boto3-oriented guidance, plus practical install and usage notes.

Security Audit

Favorites 0GitHub 0

configuring-microsegmentation-for-zero-trust

by mukul975

The configuring-microsegmentation-for-zero-trust skill helps design and validate least-privilege workload-to-workload policies for zero trust environments. Use this guide to segment applications, reduce lateral movement, and turn observed traffic into enforceable rules for Security Audit and operations.

Security Audit

Favorites 0GitHub 0

security-scan

by affaan-m

The security-scan skill audits your Claude Code .claude/ configuration for secrets, risky MCP setup, injection-prone instructions, dangerous bypass flags, and weak agent or hook definitions using AgentShield. Use it for repeatable security checks before committing or onboarding.

Security Audit

Favorites 0GitHub 156.3k

laravel-security

by affaan-m

The laravel-security skill is a practical Laravel security checklist for authn/authz, validation, CSRF, mass assignment, file uploads, secrets, rate limiting, and secure deployment. Use it for audits, feature reviews, and hardening work in Laravel apps.

Security Audit

Favorites 0GitHub 156.2k

k8s-security-policies

by wshobson

k8s-security-policies helps teams draft Kubernetes NetworkPolicy, Pod Security Standards labels, and RBAC patterns using repo-backed templates and references for hardening and audit-ready rollout planning.

Security Audit

Favorites 0GitHub 32.6k

auth-implementation-patterns

by wshobson

auth-implementation-patterns is a practical skill for designing and implementing authentication and authorization patterns, including sessions, JWT, OAuth2/OIDC, RBAC, and access control checks for APIs and apps.

Access Control

Favorites 0GitHub 32.6k

security-and-hardening

by addyosmani

The security-and-hardening skill helps harden application code before release. Use it for user input, auth, sessions, sensitive data, file uploads, webhooks, and external services, with concrete checks like input validation, parameterized queries, output encoding, secure cookies, HTTPS, and secrets handling.

Security Audit

Favorites 0GitHub 18.7k

exploiting-kerberoasting-with-impacket

by mukul975

exploiting-kerberoasting-with-impacket helps authorized testers plan Kerberoasting with Impacket GetUserSPNs.py, from SPN enumeration to TGS ticket extraction, offline cracking, and detection-aware reporting. Use this exploiting-kerberoasting-with-impacket guide for penetration testing workflows with clear install and usage context.

Penetration Testing

Favorites 0GitHub 6.2k

exploiting-idor-vulnerabilities

by mukul975

exploiting-idor-vulnerabilities helps authorized security audits test Insecure Direct Object Reference flaws across APIs, web apps, and multi-tenant systems with cross-session checks, object mapping, and read/write verification.

Security Audit

Favorites 0GitHub 6.2k

detecting-azure-service-principal-abuse

by mukul975

detecting-azure-service-principal-abuse helps detect, investigate, and document suspicious Microsoft Entra ID service principal activity in Azure. Use it for Security Audit, cloud incident response, and threat hunting to review credential changes, admin consent abuse, role assignments, ownership paths, and sign-in anomalies.

Security Audit

Favorites 0GitHub 6.1k

detecting-aws-iam-privilege-escalation

by mukul975

detecting-aws-iam-privilege-escalation helps audit AWS IAM for privilege escalation paths using boto3 and Cloudsplaining-style analysis. Use it to identify dangerous permission combinations, least-privilege violations, and security audit findings before they become incidents.

Security Audit

Favorites 0GitHub 6.1k