Security audit
Lark CLI (xieyongliang)
Security checks for vulnerabilities and agentic risk
Overview
The plugin is coherent with its description: it exposes lark-cli to agents and requires that lark-cli be installed and authenticated manually; nothing requested or installed is disproportionate to that purpose.
This plugin is coherent: it simply runs your locally installed lark-cli on behalf of an agent. Before installing, ensure you: (1) have installed and manually authenticated lark-cli (the plugin does not perform interactive auth); (2) understand that the agent can run any lark-cli command as your user (send mail, messages, access drive/base/sheets, etc.), so only install if you trust agents that will invoke it; (3) consider restricting agent permissions or testing with a limited account if possible; (4) note an implementation detail: it launches lark-cli via a shell (child_process.exec), which is normal for CLI wrappers but means malformed or unescaped input could cause unexpected shell behavior — this is an implementation risk rather than an incoherence. If you want extra safety, review index.ts (small file) and set the plugin's binaryPath or LARK_CLI_BIN to point to a vetted binary before enabling autonomous agent runs.
SkillSpector was not run because this plugin release contains no bundled skills.
Endor dependency reachability
Checked Nov 7, 2026 · 4:36 AM UTC
No reachable vulnerabilities reported
Endor dependency scan completed.
