</>MCP Agents Market
Plugin

stop-that-shit

by lennney1.7kJavaScriptUpdated 2026-09-07

Stop That Shit（别再造史了）｜面向 Codex/GPT 场景的多平台 Hook + Skill Guard：拦截 AI coding agent 无需求的哈希、校验和与任务范围膨胀。 A multi-platform Hook + Skill Guard for AI coding agents in Codex/GPT workflows: stop unrequested hashes, checksums, and task-scope creep.

CodexClaude CodeOpenCodeHermes Agent CLI

Stop That Shit is a multi-platform guard plugin for AI coding agents that intercepts and blocks unrequested actions like generating unnecessary checksums, scope creep, and task boundary violations. It works with Codex, Claude Code, OpenCode, and Hermes Agent CLI to enforce explicit user intent through hook-based validation and executable boundaries. Developers set task modes (review, change, monitor) and optional constraints (file locks, dependency permissions, hash operations) to prevent agents from performing defensive or hypothetical work not required by the current task. The plugin combines advisory skills with runtime guards that return permission denials when agents attempt actions outside defined scope.

Key Features

Multi-platform support for Codex, Claude Code, OpenCode, and Hermes Agent CLI with unified guard logic
Hook-based interception at pre-tool-use events to block unauthorized file mutations, dependency additions, and hash operations
Task mode contracts (review, change, answer, monitor, watch) that define allowed actions per session
Granular permission controls: file path locks, dependency installation flags, hash operation gates, and sub-agent budget limits
SHIT taxonomy: blocks Scope creep, unnecessary Hashing/hardening, Intent violations, and Task thrashing
Local metadata-only runtime that distinguishes checked actions, permission denials, and unobserved host effects
Paired case evaluation system with bad/good examples to refine guard boundaries
Skill-only installation mode for advisory guidance without runtime enforcement

Use Cases

  • 01Preventing AI agents from generating SHA-256 checksums or validation files when no consumer code exists
  • 02Blocking agents from modifying code during review-only tasks when user explicitly requested read-only analysis
  • 03Stopping scope creep when fixing a single bug triggers unnecessary refactoring across multiple modules
  • 04Enforcing file-path boundaries to ensure agents only modify explicitly authorized configuration or test files
  • 05Controlling dependency bloat by requiring explicit permission before agents add new npm or pip packages
  • 06Limiting sub-agent spawning budget to prevent recursive agent calls that waste tokens on redundant validation

Related Plugins

View more

andrej-karpathy-skills

A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.

Plugin216.2k

claude-mem

Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + More

Plugin95.1k

Understand-Anything

Graphs that teach > graphs that impress. Turn any code into an interactive knowledge graph you can explore, search, and ask questions about. Works with Claude Code, Codex, Cursor, Copilot, Gemini CLI, and more.

Plugin85k

rtk

CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies

Plugin82.2k

stop-that-shit — FAQ

What is Stop That Shit and which AI agents does it work with?+

Stop That Shit is a guard plugin that prevents AI coding agents from performing unrequested defensive work, scope creep, and task boundary violations. It supports Codex, Claude Code, OpenCode, and Hermes Agent CLI through platform-specific adapters that share the same guard logic.

How do I install Stop That Shit for Claude Code?+

Extract the plugin, navigate to the repo root, then run 'claude plugin validate .', 'claude plugin marketplace add ./', and 'claude plugin install stop-that-shit@stop-that-shit'. Restart Claude Code or run '/reload-plugins' to activate.

How do I install Stop That Shit for Codex?+

Run 'codex plugin marketplace add lennney/stop-that-shit' then 'codex plugin add stop-that-shit@stop-that-shit'. Restart Codex, type '/hooks' in the TUI, and trust the UserPromptSubmit and PreToolUse hooks after reviewing the commands.

What are the prerequisites and is Stop That Shit free to use?+

You need Node.js 18+ and one of the supported AI coding agents (Codex, Claude Code, OpenCode, or Hermes Agent CLI). The plugin is open source under the MIT license and free to use with no API keys required.

How do I set task boundaries with Stop That Shit?+

Use command prefixes like '$stop-that-shit review' for read-only tasks or '$stop-that-shit change' for modifications. Add constraints like 'files=src/config.cjs' to lock file paths, 'deps=allow' to permit dependency changes, 'hash=allow' for checksum operations, or 'agents=1' to limit sub-agent budget.

Can I use Stop That Shit as an advisory skill without runtime hooks?+

Yes, you can install only the skill file for guidance without executable enforcement. For Claude Code, copy SKILL.md to ~/.claude/skills/stop-that-shit/; for Codex, use the skill-installer to fetch from the GitHub repo. Skill-only mode provides task mode prompts but won't block unauthorized actions.

How do I install stop-that-shit?+

Open the source repository on GitHub and follow its README. stop-that-shit is a plugin — MCP Agents Market links you directly to the official repo.

Is stop-that-shit free?+

stop-that-shit is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.

Related searches

Developer Tools PluginsCode Review PluginsProductivity & Workflow PluginsBest MCP ServersBest SkillsBest AgentsTop rated listingsClaude Code toolsBrowse all categories