registry / @rigxyz/cli / 0.10.2

@rigxyz/cli@0.10.2

Portable, runnable agent workspaces

AI Security Review

scanned 16d ago · by lpm-firewall-ai

LPM flags this version as an AI-agent control-surface risk. Installation mutates the user's global Claude Code skill directory. The injected instructions influence agent behavior in every folder and can direct collaboration actions.

Static reason

One or more suspicious static signals were detected.; previous stored version diff introduced dangerous source

Trigger

npm postinstall when ~/.claude exists and RIG_SKIP_SKILL_INSTALL is unset.

Impact

Broad AI-agent control-surface hijack and potential unprompted agent actions.

Mechanism

Unconsented global AI-agent skill installation.

Policy narrative

On installation, the package silently creates ~/.claude/skills/rig/SKILL.md. That user-level Claude Code skill declares broad triggers beyond explicit rig use and instructs the agent to run rig/tapd commands and perform collaboration behavior. This is an unconsented postinstall mutation of a foreign, global AI-agent control surface.

Rationale

Source inspection confirms a postinstall hook writes a persistent global Claude Code instruction file. The guards and opt-out do not establish user consent for this broad control-surface modification.

Evidence

package.jsonbin/postinstall.mjs~/.claude/skills/rig/SKILL.md

Decision evidence

public snapshot

AI identified this as a dangerous AI-agent capability at 99.0% confidence with low false-positive risk.

@rigxyz/cli@0.10.2

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Source & flagged code