clean_claude_directory

THE RISK

Critical Risk

Why clean_claude_directory needs a policy

The tool irreversibly deletes multiple types of data (debug logs, todos, snapshots, orphaned data) from the .claude directory. These deletions cannot be undone, making this Destructive. The blast radius is high because it could remove important snapshots, conversation history, or configuration data that a user may need for recovery or auditing purposes.

From the tool's definition Removes debug logs, empty todos, old snapshots, and orphaned data to free disk space

Documented attack patterns abuse exactly the kind of access clean_claude_directory gives an agent:

POLICY

How to control clean_claude_directory

PolicyLayer is an MCP gateway — it sits between your AI agents and Claude Code Toolkit, and nothing reaches the server without passing your rules. This is the rule we recommend for clean_claude_directory:

policy.json

{
  "version": "1",
  "default": "deny",
  "hide": [
    "clean_claude_directory"
  ]
}

clean_claude_directory disappears from the agent's tool list entirely, and any attempt to call it is denied. The rest of the server keeps working.

Create a free account and register Claude Code Toolkit — nothing to install.
Add this policy — paste it, or build it visually.
Point your MCP client (Claude, Cursor, anything) at your gateway URL.

RESTRICT THIS TOOL →

Free to start. No card required.

EXPLORE

Related tools and policies

More Claude Code Toolkit tools

Destructive cleanup_backups Delete old backup files to free up disk space. Destructive wipe_traces Securely wipe ALL Claude Code traces. Overwrites files with zeros before deletion. Require Destructive clean_traces Selectively clean Claude Code traces by category, age, or project. Destructive enforce_retention Apply data retention policy by deleting sessions older than specified days. Execute run_doctor Full integrity check: detects corrupted JSONL lines, truncated thinking blocks, oversized Execute run_maintenance Run maintenance checks on Claude Code installation. Identifies issues, old backups, and ar Execute start_dashboard Start the web dashboard on a local port. Returns the URL to open in a browser. Execute unstick_session Unstick a Claude Code session that

All 40 Claude Code Toolkit tools →

Destructive tools on other servers

Android Forensics ADB MCP Server create_full_backup 0xarchive web3_revoke_key Frida Game Hacking MCP clear_scan cPanel MCP Server clear_spam_box

Go deeper

Critical-risk tools in Claude Code Toolkit → Every critical-risk tool this server exposes, in one view.
All critical-risk MCP tools → Every critical-risk tool across the catalogue, and the policy pattern that contains them.
The MCP Attack Database → Documented attack patterns against MCP deployments — and the policies that stop them.
Blast radius → How much damage one compromised or confused agent can do — and how to bound it.
Tool call approval → Approval gates on individual tool calls: auto-allow, prompt, or deny.
Rate Limiting MCP Tool Calls: A Practical Guide →
MCP Security: Why Prompt Guardrails Aren't Enough →

FAQ

Questions about clean_claude_directory

What does the clean_claude_directory tool do? +

Analyze and clean the .claude directory. Removes debug logs, empty todos, old snapshots, and orphaned data to free disk space. It is categorised as a Destructive tool in the Claude Code Toolkit MCP Server, which means it can permanently delete or destroy data. Block by default and require explicit approval.

How do I enforce a policy on clean_claude_directory? +

Register the Claude Code Toolkit MCP server in PolicyLayer and add a rule for clean_claude_directory: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Claude Code Toolkit. Nothing to install.

What risk level is clean_claude_directory? +

clean_claude_directory is a Destructive tool with critical risk. Critical-risk tools should be blocked by default and only enabled with explicit human approval.

Can I rate-limit clean_claude_directory? +

Yes. Add a rate_limit block to the clean_claude_directory rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.

How do I block clean_claude_directory completely? +

Set action: deny in the PolicyLayer policy for clean_claude_directory. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.

What MCP server provides clean_claude_directory? +

clean_claude_directory is provided by the Claude Code Toolkit MCP server (asifkibria/claude-code-toolkit). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.

Enforce policy on every Claude Code Toolkit tool call.

Start from Claude Code Toolkit, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.

CHECK YOUR STACK →