registry / dingtalk-workspace-cli / 1.0.59

dingtalk-workspace-cli@1.0.59

DingTalk Workspace CLI - AI-powered productivity tools

AI Security Review

scanned 12h ago · by lpm-firewall-ai

LPM flags this version as an AI-agent control-surface risk. Installing the package automatically deploys bundled AI skills into the user's canonical agent directory and detected third-party agent skill directories. The installed skills become available after agent restart.

Static reason

High-risk behavior combination matched malicious policy.

Trigger

npm postinstall

Impact

Unconsented modification of multiple AI-agent instruction/capability surfaces.

Mechanism

extracts bundled skills and publishes them into broad AI-agent skill control surfaces

Policy narrative

On npm installation, package.json invokes install.js. It extracts a bundled skill archive, installs it by default to ~/.agents/skills, detects numerous unrelated AI-agent installations, and publishes symlinks or copied skills into their skill directories. This changes behavior-bearing AI-agent control surfaces without an explicit user setup command; existing managed skill paths can be backed up and replaced.

Rationale

This is concrete, install-time mutation of foreign and broad AI-agent control surfaces. The absence of network exfiltration does not mitigate that policy-triggering behavior.

Evidence

package.jsoninstall.jsassets/dws-skills.zipvendor/dwsshare/skills~/.agents/skills~/.claude/skills~/.codex/skills~/.cursor/skills

dingtalk-workspace-cli@1.0.59

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Source & flagged code