registry / @vanillagreen/pi-questions / 2.0.4

@vanillagreen/pi-questions@2.0.4

Pi extension for structured multi-tab inline questions and bridge-driven replies.

AI Security Review

scanned 3d ago · by lpm-firewall-ai

LPM treats this as warn-only first-party agent extension lifecycle risk. Automatic installation modifies Pi system prompt content through a package-marked instruction block. This creates a guarded extension lifecycle risk; inspected behavior does not establish a malicious attack.

Static reason

One or more suspicious static signals were detected.

Trigger

npm postinstall, with removal through preuninstall.

Impact

Question-tool guidance becomes persistent agent context without a separate setup command; other prompt content is preserved.

Mechanism

The lifecycle helper locates a Pi scope and appends package instructions to APPEND_SYSTEM.md, including an existing user-directory fallback.

AI rationale

Source inspection supports a guarded platform extension lifecycle warning, without confirmed malicious behavior. Citation issuance was attempted but the package tool transport was closed; source inspection succeeded through read-only filesystem commands.

Evidence

package.jsonscripts/append-system.mjsinstructions.mdAPPEND_SYSTEM.md

Decision evidence

public snapshot

AI called this Suspicious at 85.0% confidence as Dangerous Capability with medium false-positive risk.

Evidence for warning

package.json automatically runs scripts/append-system.mjs on postinstall and preuninstall.

@vanillagreen/pi-questions@2.0.4

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Affected versions and remediation

Source & flagged code