claude-jev-plugin
Claude Code plugin adding Jev judgments to Bash, Write and Edit calls and their outputs; a semantic guardrail, not a sandbox.
Overview
This claude-jev plugin by dr-dimitru adds TypeSafe Jev semantic judgments to Claude Code tool execution. PreToolUse judges Bash, Write and Edit calls; PostToolUse judges successful Bash output; PostToolUseFailure judges failed Bash output. Requests go directly from local plugin code to TypeSafe's HTTPS API, and there is no MCP server.
The README is explicit that it is a semantic guardrail, not a security sandbox: TypeSafe can be unavailable or wrong, and Claude Code permission rules remain the hard control. The plugin is disabled by default because enabling it sends bounded data to an external service and may cost money, and the key is only sent as a bearer token over HTTPS (never to a local endpoint). It is BSD-3-Clause.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Yes / no | Is this Bash, Write or Edit call risky?(inferred) | yes · no |
| Yes / no | Does this command output indicate a problem?(inferred) | yes · no |
How it works
- Claude proposes a Bash, Write or Edit call
- PreToolUse sends bounded data to Jev
- Plugin acts on the judgment
- Post-tool hooks judge command output
Strengths
- Disabled by default with clear privacy notes
- No MCP server; direct HTTPS
- Careful key handling
Caveats
- Not a sandbox; permission rules stay the hard control
- Sends data to TypeSafe and may incur cost
Getting started
/plugin marketplace add dr-dimitru/claude-plugins-marketplace, /plugin install claude-jev@dr-dimitru-claude-tools --scope user, export TYPESAFE_API_KEY, then enable it.At a glance
- Type
- Plugin
- Status
- beta
- Runtime
- Claude Code plugin
- Has code
- yes
- License
- BSD-3-Clause
Use cases and integrations
How curated lists describe it
TypeSafe Jev semantic guardrails for Claude Code
logicrw/awesome-jev-projectsTypeSafe Jev semantic guardrails for Claude Code.
hellogumbo/awesome-jev
Frequently asked questions
Is it a sandbox?
No, it is a semantic guardrail; Claude Code permissions remain the hard control.
Is it on by default?
No, it ships disabled.
Which tools are judged?
Bash, Write and Edit calls, plus Bash outputs.
Related projects
Local guard layer for coding agents that blocks dangerous tool calls by deterministic rules, with optional Jev judgments.
CLIbetaTypeScriptAgent-ergonomic Jev CLI and PreToolUse hook that scores shell commands for hazards and screens fetched text for injection.
PluginbetaTypeScriptPreToolUse firewall for Claude Code and Codex: shell-aware rules first, then Jev returns allow, ask or block, failing closed.
HookbetaTypeScriptSafety gate plugin for Claude Code, Codex CLI and Antigravity: local sensitive-file checks, then Jev judges blast radius.
PluginbetaTypeScriptSecurity hook giving Codex, Copilot, Gemini, Cursor, pi, OpenCode and ACP an auto-mode-like guard via three Jev questions.
LibrarybetaTypeScriptClaude Code mod that answers auto mode's tool-call permission question with eight Jev yes/no checks behind a shell-aware blocklist.
ProjectbetaTypeScript