registry / @maxiaochao/pi-toolkit / 0.1.5

@maxiaochao/pi-toolkit@0.1.5

Personal pi coding agent toolkit: /todos extension + /cache_export interactive cache dashboard.

AI Security Review

scanned 3h ago · by lpm-firewall-ai

LPM flags this version as an AI-agent control-surface risk. Installing the package changes a global Pi agent control file in the user's home directory. The write is automatic and replaces existing content when it differs.

Static reason

High-risk behavior combination matched malicious policy.

Trigger

npm installation runs the postinstall hook.

Impact

The package can silently alter instructions used by the user's Pi agent across projects.

Mechanism

Automatic overwrite of global agent instructions.

Policy narrative

The manifest automatically executes a script after installation. That script reads the package's bundled instructions, creates the global Pi agent directory if necessary, and overwrites the user's existing global agent instructions whenever they differ. This is an unconsented install-time mutation of a broad AI-agent control surface.

Rationale

The postinstall hook silently overwrites a global AI-agent instruction file outside the package and consumer project. This meets the policy for malicious install-time AI-agent control hijacking.

Evidence

package.jsonscripts/install-agents.mjsglobal/AGENTS.md~/.pi/agent/AGENTS.md

Decision evidence

public snapshot

AI identified this as a dangerous AI-agent capability at 97.0% confidence with low false-positive risk.

@maxiaochao/pi-toolkit@0.1.5

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Source & flagged code