Auto-approve compound Bash commands in Claude Code — pipes, chains, subshells parsed and checked per-segment
-
Updated
Jul 29, 2026 - Shell
#
pretooluse
Here are 28 public repositories matching this topic...
Moved to cc-safe-setup (910 hooks). This repo keeps the original 16 hooks for backward compatibility.
plugin bash open-source hooks npm security devops automation cookbook owasp developer-tools safety autonomous ai-safety claude anthropic claude-code database-protection pretooluse posttooluse
-
Updated
Aug 8, 2026 - Shell
🐕 AI-Powered Risk Assessment for Claude Code — 7-layer pipeline, native notifications, menu bar dashboard. One line install.
notifications macos rust cli open-source hooks security cross-platform notifier developer-tools risk-assessment claude menu-bar ai-security ai-agent anthropic claude-code claude-code-hooks pretooluse claude-code-notifier
-
Updated
Apr 21, 2026 - Rust
Flexible and deterministic Claude Code PreToolUse handler
-
Updated
Jul 2, 2026 - Haskell
Runtime controls for Claude Code derived from real tool-call telemetry. Pair with cc-logger to observe failures, promote recurring patterns into rules, nudge recoverable mistakes, and block irreversible ones.
hooks observability ai-agents guardrails runtime-controls llmops claude-code agent-safety agent-governance pretooluse
-
Updated
Aug 18, 2026 - Python
Branch-aware git permissions for Claude Code — auto-approve safe git/gh commands, pause at main and destructive ops.
git bash productivity hooks security permissions developer-tools gh worktree guardrail claude-code claude-code-plugin pretooluse
-
Updated
Aug 21, 2026 - Shell
A PreToolUse Bash tool hook that auto-approves read-only commands
-
Updated
Jul 15, 2026 - Go
An environment-aware Claude Code PreToolUse hook for the Railway CLI. Resolves the target environment at call time - from -e or the per-directory link - so production deploys, volume deletes, and live DB shells hard-deny while read-only troubleshooting stays prompt-free. Fails closed.
-
Updated
Jun 25, 2026 - Shell
The layer a sandbox doesn't give you: a PreToolUse hook that blocks unapproved spending, account creation and fund movement before the call runs, queues them for human review, and proves it actually ran -- including whether your harness fired it for subagent tool calls. Plus a spend ceiling and loop detector.
python hooks security budget cost-control autonomous-agents ai-safety human-in-the-loop ai-agents guardrails approval-workflow llm llmops anthropic token-usage llm-agents agent-tools claude-code agent-safety pretooluse
-
Updated
Aug 22, 2026 - Python
Block irreversible AI-agent actions before they run — deterministic, fail-closed action veto for Claude Code, Codex, Cursor, and other tool-using agents.
python security mcp developer-tools cursor codex ai-safety devsecops ai-agents guardrails llm local-llm ollama llm-agents claude-code fail-closed agent-safety pretooluse agent-guardrail action-veto
-
Updated
Aug 22, 2026 - Python
Deterministic, LLM-free PreToolUse guardrail for Claude Code and Codex CLI. Applies local policy to shell commands and supported file operations, with deny/ask decisions and audit logs.
cli golang hooks developer-tools audit-log deterministic policy-engine devsecops security-tool guardrails llm-security file-access-control claude-code codex-cli ai-coding-agent ai-agent-security pretooluse
-
Updated
Aug 22, 2026 - Go
AI Agent Memory — Fewer prompts, smarter decisions. Remembers your trust decisions across Claude Code, Codex & MCP. Free & open source.
open-source hooks security typescript mcp codex ai-security ai-agent ai-memory ai-guardrails claude-code pretooluse confirmation-fatigue
-
Updated
Aug 17, 2026 - TypeScript
Compile your CLAUDE.md rules to real Claude Code hooks.
-
Updated
Aug 1, 2026 - TypeScript
Claude Code-compatible command hooks for the OMP (Oh My Pi) coding agent
-
Updated
Jul 14, 2026 - TypeScript
Recoverability-first runtime controls for the OpenAI Codex CLI: learn from real tool-call telemetry, nudge repeated mistakes, and block irreversible actions before they run.
-
Updated
Aug 18, 2026 - Python
One-file PreToolUse hook that stops Claude Code from running catastrophic Bash commands (rm -rf ~, force-push to main, curl|sh). Zero-dep, fails open, 38-case tested.
-
Updated
Jul 8, 2026 - Python
Guard rails for Claude Code: workspace boundaries, protected branches, production targets, lost exit status, and foreground time.
git shell bash kubernetes hooks security terraform sandbox permissions developer-tools ai-agents guardrails llm anthropic agentic-ai claude-code claude-code-hooks claude-code-plugin claude-code-marketplace pretooluse
-
Updated
Aug 22, 2026 - Python
A PreToolUse hook for Kiro that allows harmless tool calls, asks about sensitive ones and blocks destructive ones.
-
Updated
Aug 20, 2026 - JavaScript
Zero-dependency Claude Code PreToolUse hook that blocks reading credential files into agent context, without false-positives on mere mentions.
-
Updated
Jul 31, 2026 - JavaScript
Road signs for coding agents: one measured fact at the moment of action, silence otherwise
python git developer-tools ai-agents git-worktree agentic coding-agents claude-code agent-observability agent-reliability context-engineering claude-code-hooks claude-code-plugin agent-harness context-integrity pretooluse zero-llm hermes-labs
-
Updated
Aug 11, 2026 - Python
Improve this page
Add a description, image, and links to the pretooluse topic page so that developers can more easily learn about it.
Add this topic to your repo
To associate your repository with the pretooluse topic, visit your repo's landing page and select "manage topics."