Clauder workflow
A Claude Code configuration with safeguards: approvals for risky actions, command audit, sub-agents and 65 on-demand MCP servers.
- 59 ★ on the repository
- Last commit
- Apache-2.0
Broad access
Why this level
- The installer changes the Claude Code configuration and installs MCP servers, and the log records every shell command and MCP call.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add workflows/spacehendrix-clauderDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- The installer changes the Claude Code configuration and installs MCP servers, and the log records every shell command and MCP call.
This is third-party code. Review the repository files before installing.
What it does
Clauder drops a set of Claude Code settings into a project. Protection includes immutable files and ignore patterns, human approval of sensitive operations, limits on destructive git operations, environment variable protection and a `/guardrails` command for per-project access rules. An optional layered secret detection can be switched on. Audit runs on a SQLite database: shell commands and MCP calls are logged and a web app shows them live. Git checkpoints are made before and after each interaction. The set includes sub-agents for investigation, review, testing, security review and documentation, 67 specialized agents in expansion packs across eight domains and over 65 MCP servers installed on demand. Auto-update preserves your configuration.
Who it is for. Developers and teams who run Claude Code in working repositories and want guardrails, an action log and a ready set of agents.
Good fit when
- The agent must be barred from sensitive files and destructive git operations.
- A team wants a log of commands and MCP calls with a live view.
- MCP servers are added to a project as needed with no manual config editing.
- Ready sub-agents for review, testing and security checks are needed.
Not a fit when
- You already have your own `.claude` configuration that must not be overwritten.
- There is no git repository: the project must be one.
- You expect an autonomous safety guarantee with no human supervision.
Example request
Run /guardrails for this project and list which files and operations now require my approval.Limitations
Installation may overwrite an existing `.claude` configuration: the author makes backups in `.claude-backup/` and `.claude-mcp-backup/`. Python 3.10 or newer, `pip`, `git`, `jq` and `claude` are required, plus `npx`, `uvx`, `uv`, `pipx`, `pnpx` or `docker` for MCP servers. It runs in `bash` and `zsh`, on Windows only from `bash`, and some MCP servers need manual setup. The author warns outright that the guardrails are insufficient without human supervision.
How to disable. Restore the previous setup from `.claude-backup/` and `.claude-mcp-backup/`, remove the `.claude` and `.mcp.json` created by Clauder and delete the repository clone.
FAQ
How do I install Clauder?
Run `cd ~/ && git clone https://github.com/spacehendrix/clauder.git && cd clauder && bash ./clauder_install.sh`. The project you work in must be a git repository.
What does the /guardrails command do?
It sets access rules for the current project: what the agent may do, what needs approval and what is closed.
Will Clauder overwrite my configuration?
It may. Previous files are copied to `.claude-backup/` and MCP settings to `.claude-mcp-backup/`, from where they can be restored.
Which agents ship in the expansion packs?
67 specialized agents across eight domains: frontend, backend, data, AI, infrastructure, games, desktop and general software.
How can I see what the agent did?
Shell commands and MCP calls are written to a SQLite database, and a web app shows the log in real time.
Discussion
No questions or reviews yet.
Related
Workflowsclaude-mods adds 112 skills, 3 agents and 3 commands to Claude Code: stacks, agent operation, defense, session continuity.
A personal ~/.claude for Claude Code: twelve subagents, review commands, worktree and goal skills and a Go status line.
Claudette runs several Claude Code sessions in parallel, each in its own git worktree, from a window or the console
CC_GodMode installs a self-orchestrating team of 14 agents and 14 skills into Claude Code with hooks and quality gates.
MCP server of the self-hosted EmperorClaw panel: Claude Code and Codex read company knowledge, tasks and customers in scope
A set of /sc slash commands, agents and behavior modes for Claude Code, from brainstorming and research to implementation and tests
README in short
The author presents Clauder as a safety-first toolkit for Claude Code with auto-update, extensible to any project. It covers protection, logging, automation, sub-agents, domain expansion packs and installation.