registry / shmakk / 1.2.5

shmakk@1.2.5

AI-supervised terminal wrapper — command correction, tool-driven tasks, safety controls

AI Security Review

scanned 4h ago · by lpm-firewall-ai

LPM treats this as warn-only first-party agent extension lifecycle risk. Install-time setup mutates the package-owned shmakk AI-skill directory. No confirmed exfiltration, remote payload execution, or foreign agent-control-surface modification was found.

Static reason

One or more suspicious static signals were detected.

Trigger

npm install executes postinstall.

Impact

Bundled instructions become globally available to this package's AI-agent runtime.

Mechanism

Copies bundled agent skill Markdown into ~/.config/shmakk/skills.

Rationale

This is a real install-time agent-extension lifecycle risk, not concrete malware. The package’s own global skill state is modified without an explicit runtime command, so it warrants a warning under the stated policy.

Evidence

package.jsonscripts/install-skills.jsscripts/patch-onnxruntime.jssrc/skills.jssrc/tools.jssrc/safety.jsskills/

Decision evidence

public snapshot

AI called this Suspicious at 92.0% confidence as Dangerous Capability with low false-positive risk.

Evidence for warning

package.json runs postinstall.

shmakk@1.2.5

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Source & flagged code