AI Coding Token Economizer & Model Fatigue Shield for Cursor, Claude Code, Windsurf & VS Code
Soft block: AI-agent control surface
Warn by default; block when configured. Agent clients may launch the package's MCP server and invoke its registered tools automatically.
AI security review
scanned · by lpm-firewall-ai
The postinstall performs unconsented writes to global AI-agent control configuration and registers an automatically allowed MCP server. This meets the control-surface hijack block policy.
Agent clients may launch the package's MCP server and invoke its registered tools automatically.
Mechanism
The hook calls self-install, which writes a TokenShield entry to global MCP configuration files without an affirmative consent check.
Policy narrative
During npm installation, the silent postinstall invokes self-install. That command performs global MCP registration even in node_modules and writes a TokenShield server entry into agent configuration files without asking the installer. The entry runs this package's CLI as an MCP server and lists tools as automatically allowed, changing how global agent clients can interact with the package.
Avoid installing tokenshield-contextplus@1.1.3. Remove it from direct dependencies and check your lockfile for transitive copies.
Choose an independently verified alternative or release. This report does not establish that other versions are safe.
If this version ran, check your AI agents' configuration and instruction files (the files this report lists) for changes you did not make, and remove them. This finding does not by itself establish malware intent.