registry / psypi / 0.3.2

psypi@0.3.2

A standalone, evidence-grounded social-science research agent — a conversation-driven agent that bundles the Pi runtime with psypi's research contracts, evidence gates, and workflows

AI Security Review

scanned 3h ago · by lpm-firewall-ai

LPM flags this version as an AI-agent control-surface risk. Installing the package triggers an unprompted rewrite of a separately owned Pi coding-agent dependency. The mutation changes its package metadata and interactive-mode code.

Static reason

One or more suspicious static signals were detected.

Trigger

npm postinstall during psypi installation

Impact

Changes the installed Pi agent control/runtime surface without explicit user approval.

Mechanism

postinstall patches a foreign AI-agent package

Policy narrative

On installation, npm invokes scripts/rebrand-pi.mjs. It resolves @earendil-works/pi-coding-agent, modifies that package's package.json, and rewrites its interactive-mode JavaScript. This is an unconsented install-time mutation of a foreign AI-agent runtime, even though the observed edits are branding-oriented and no exfiltration was found.

Rationale

The package has a concrete unconsented postinstall chain that rewrites a foreign AI-agent dependency. That meets the install-control-surface block boundary.

Evidence

package.jsonscripts/rebrand-pi.mjs@earendil-works/pi-coding-agent/package.json@earendil-works/pi-coding-agent/dist/modes/interactive/interactive-mode.js

Decision evidence

public snapshot

AI identified this as a dangerous AI-agent capability at 98.0% confidence with false-positive risk.

psypi@0.3.2

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Source & flagged code