Safety plugins

PreToolUse guard for Bash: blocks dangerous shell commands (rm -rf ~, fork bombs, curl|sh, force push to main) before they run, with three safety levels and opt-in per-level ask mode.

hooks

Branch-aware git guardrails plus destructive gh CLI protection: denies pushes to main/master, protected-branch deletion, direct commit/merge/rebase/reset/push on a protected branch, and gh pr merge/close, issue close, release and repo delete.

hooks

Stops "fake green" test-gaming: blocks the agent from deleting, renaming-away, or skip/xfail-disabling tests so it fixes the code instead, across pytest, jest, Go, JUnit, RSpec, and more.

hooks

Blocks unsafe git subtree invocations (URL fetch, missing --squash/-S, subtree push)

hooks

Personal agent operating system for Claude Code. Safety hooks, memory tiers, agents, commands, and cross-engine hard enforcement.

200 skills

Personal agent operating system for Claude Code. Safety hooks, memory tiers, agents, commands, and cross-engine hard enforcement.

200 skills

5 essential safety hooks: block rm -rf, force-push, hard-reset, env overwrites, and package publish. The minimum viable safety net.

hooks

Git safety hooks: block force-push, protect main/master, prevent hard reset, guard interactive rebase, and block git clean -fd.

hooks

PreToolUse hook that denies git commands staging protected paths, whole-tree adds, or rewriting history.

hooks

PreToolUse hook that blocks IaC write/destructive commands (ansible-playbook/ansible-pull without --check, ad-hoc ansible write modules, terragrunt/tofu/terraform apply-family including mise-wrapped forms), so agents never apply changes to real infrastructure.

hooks

Mandatory confirmations for destructive operations, migration and database safety, file and Git safety. 5 skills (same as the safety entry).

5 skills

A seatbelt for YOLO mode: blocks catastrophic shell commands (rm -rf ~, mkfs, dd to a disk, fork bombs) and asks before risky ones (force-push, reset --hard, DROP TABLE, curl | sh), even with permissions bypassed.

Keeps Claude's file edits inside the project: refuses writes outside it (symlinks resolved), never lets a tool touch .git internals, and asks before it changes lockfiles, CI workflows, migrations or secrets.

Keeps secrets out of Claude's context: API keys, tokens, private keys and passwords in tool output are replaced with [REDACTED:kind] before the model or the transcript ever sees them.

Fetch and sanitize web content with URL/IP checks while preserving its untrusted-source status.

1 skill · 3 commands · 1 agent · 1 MCP server · hooks

Score a CRM-connected agent's permissions and approval cards. Refuses a CRM write with no prompt.

1 skill

Config-driven file/directory guard — lock blocks pair protected paths with an unlock phrase and a PreToolUse hook blocks Edit/Write/Bash/Task mutations unless the phrase is spoken; also conceals paths (hard read+write deny, no unlock) and hardens MCP secrets behind a setuid OS boundary

9 skills · hooks

No AI self-attribution anywhere a change gets signed — PreToolUse hooks block co-author trailers, generated-with footers, attribution links and session trailers in commit messages, pull request text and the files themselves, plus AI or absent committer identities; opt-in git hooks cover what happens outside the session; bare product mentions always ship

2 skills · hooks

5 essential safety hooks: block rm -rf, force-push, hard-reset, env overwrites, and package publish. The minimum viable safety net.

hooks

Git safety hooks: block force-push, protect main/master, prevent hard-reset, guard rebase, and auto-backup before destructive operations.

hooks

Set up the guard mods for this project from its own files: the fence paths, the pins, a .claude/style-cop.md, and the red-green test command, as commands to paste and one file to write.

1 skill

Audit a Claude Code plugin, mod, or skill before you install it: what it would do on your machine, what its README doesn't say, and whether to install.

1 skill

Undo any agent turn, including shell commands and untracked files

Holds rm -rf, force pushes, git reset --hard, DROP TABLE, curl | sh, sudo and .env edits behind a question; refuses the unrecoverable ones.

Stop Claude from retrying the same failing shell command over and over. After N identical failures the next attempt is held, and you choose: stop, or try once more.

Lifecycle orchestrator, safety rails, and a founder-friendly skill library for non-technical solo founders building products with AI coding agents.

16 skills · 5 commands · 3 agents · 8 MCP servers · hooks

Automated technical and disclosure compliance-assistance audit of a project: profiling, read-only specialist subagents, fresh research, optional Codex peer review, one consolidated report. Use on /app-audit, "audit this app", or a pre-launch security/privacy/store check.

27 skills

Check what is actually live (store listing and labels, published policies, deployed backend, retention jobs) against the code and published words, then work the remediation runbook. Use on /compliance-check, "what's left on compliance", "close out the audit", or before a store submission.

27 skills

Independent fresh-context review before anything ships. Use on "ship it", "release", "submit", "deploy", "push to production", before any production push, and after a large diff, even when the diff looks fine.

27 skills

WI・委譲・承認・記録・完了報告の運用を提供するプラグイン。成果物の品質判断にはagent-basicsを使う。

14 skills · 2 MCP servers · hooks