taiwan-md-knowledge-base
Warn
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to clone a repository from an external GitHub account (
https://github.com/frank890417/taiwan-md.git). - [REMOTE_CODE_EXECUTION]: Instructions specify running
npm installandnpm run devafter cloning, which facilitates the execution of scripts and dependencies defined in the external repository. - [COMMAND_EXECUTION]: The skill requires the execution of a local shell script (
bash scripts/sync.sh) and usescurlto download image assets from Wikimedia Commons. - [PRIVILEGE_ESCALATION]: The documentation suggests using
chmod +xon thescripts/sync.shfile which is part of the downloaded repository content, potentially granting execution privileges to unverified scripts. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process Markdown files from the
knowledge/directory. This ingestion surface lacks explicit boundary markers or sanitization, creating a potential vector for indirect prompt injection if external contributors provide malicious content.
Audit Metadata