registry / @z2h-academy/tutor / 0.5.11

@z2h-academy/tutor@0.5.11

MCP server que expone los Developer Docs (MD de roadmaps) a TUIs agentic (OpenCode, Codex, Claude Code)

AI Security Review

scanned 1mo ago · by lpm-firewall-ai

Installation automatically modifies a user-wide OpenCode control surface. It registers an enabled MCP server and installs package-controlled skills.

Static reason

High-risk behavior combination matched malicious policy.

Trigger

Installing the package with npm lifecycle scripts enabled.

Impact

The package changes the agent configuration and adds instructions that OpenCode can load without an explicit setup command.

Mechanism

Postinstall registration of an MCP command and copying of agent skills.

Attack narrative

During installation, the lifecycle hook invokes modules that alter the user's OpenCode configuration and copy package-supplied skills into OpenCode's global skill directory. The configuration enables a local MCP command named tutor-mcp. These changes occur without an explicit user setup action and affect a broad AI-agent control surface.

AI rationale

The package contains an automatic postinstall path that mutates a foreign, user-wide AI-agent configuration and installs agent skills. This is concrete unconsented install-hook abuse under the stated policy.

Evidence

package.jsondist/cli/setup.jsdist/cli/install-skills.js~/.config/opencode/opencode.json~/.config/opencode/skills/<skill-name>/SKILL.md

Decision evidence

public snapshot

with false-positive risk.

@z2h-academy/tutor@0.5.11

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Affected versions and remediation

Source & flagged code