registry / @volcengine/ark-cli / 1.0.29

@volcengine/ark-cli@1.0.29

火山方舟 ARK 平台命令行工具

AI Security Review

scanned 23d ago · by lpm-firewall-ai

On installation, the package fetches and executes an opaque native binary. It then automatically asks that binary to refresh and install skills into local AI agents.

Static reason

One or more suspicious static signals were detected.; source matched previously finalized malicious package; routed for review; source fingerprint signature matched known malicious package; routed for review

Trigger

npm postinstall

Impact

Unconsented install-time mutation of AI-agent control surfaces through code absent from the package snapshot.

Mechanism

remote binary bootstrap followed by automatic agent skill installation

Attack narrative

Installing the package runs postinstall. The hook downloads a platform-specific executable from manifest-controlled URLs, verifies it, marks it executable, and runs it. It then invokes +connect --refresh without a user command; package comments state this detects agents and installs skills. Because the binary and skill logic are not included in the package, the automatic cross-agent mutation is opaque to static review.

AI rationale

This is an automatic lifecycle chain that executes a remotely obtained opaque binary and broadly installs AI-agent skills. The embedded hash does not make the unreviewable, unconsented agent-control mutation safe.

Evidence

package.jsonscripts/postinstall.jsmanifest.jsonbin/arkcli-${platform}-${arch}${ext}skills/

Network endpoints2

lf3-static.bytednsdoc.com/volcengine/ark-cli/releases/download/v1.0.29

@volcengine/ark-cli@1.0.29

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Affected versions and remediation

Source & flagged code