Kenryu42 Claude Code Safety Net
kenryu42-claude-code-safety-net is a Claude Code plugin for the Ship phase that hooks command execution to block destructive git and filesystem operations before they run.
by kenryu42 · github.com/kenryu42/claude-code-safety-net
Add a Claude Code hook that intercepts and blocks destructive git and filesystem commands before the agent can run them.
Add it to Claude Code
Install the plugin in Claude Code. One command, paste-ready.
/plugin install kenryu42-claude-code-safety-net@kenryu42/claude-code-safety-netBuilt to be called by your agent
Skillselion is itself an MCP server. Your agent can pull this entry and a paste-ready install config straight from the API - no copy-paste.
Retrieve this entry with skillselion.get_details("plugin:kenryu42/claude-code-safety-net") and the paste-ready config with skillselion.get_install_config("plugin:kenryu42/claude-code-safety-net").
What it does
kenryu42-claude-code-safety-net is a Claude Code plugin that registers a coding-agent hook to act as a safety net against destructive git and filesystem commands. Solo builders who let Claude run shell and git tools need guardrails: one mistyped rm -rf, git reset --hard, or force-push can erase hours of work or pollute main. This plugin catches those classes of commands before they execute, shifting security left into every session rather than relying on manual review after the fact. It is intentionally narrow compared to full workflow bundles—one plugin, one job—making it easy to adopt alongside other skills. Install it when you routinely grant file and git permissions to the agent during build and ship. It complements code review and testing plugins by preventing irreversible damage in the first place. Complexity is beginner-friendly: enable the hook and let it filter risky invocations automatically.
Highlights
- Hook-based safety net that runs before tool execution in Claude Code
- Blocks destructive git operations before they mutate history or remotes
- Blocks dangerous filesystem commands before data loss
- Lightweight single-plugin install focused on command interception
- Community-maintained guardrail for everyday agent sessions
Why builders use it
Autonomous agents can accidentally run irreversible git or shell commands that delete repos, branches, or local data.
After install, risky destructive commands are intercepted by the hook so sessions fail safe instead of silently damaging your project.
At a glance
- Type - Plugin in Security.
- Adoption - 0 installs, 1.2k stars, 0 votes.
FAQ
Who is kenryu42-claude-code-safety-net for?
It is for developers using Claude Code who grant the agent git and filesystem access and want automatic blocking of destructive commands.
When should I use kenryu42-claude-code-safety-net?
Enable it before any session where the agent can run shell or git tools, especially during refactors, cleanups, and release branches.
How do I add kenryu42-claude-code-safety-net to my agent?
Install the kenryu42/claude-code-safety-net plugin in Claude Code and activate its hook so command proposals are validated prior to execution.
Comments
Share how you use kenryu42-claude-code-safety-net, gotchas, or tips for other indie builders.
No comments yet - be the first to share how you use it.