find-skills
Fail
Audited by Gen Agent Trust Hub on Feb 21, 2026
Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- EXTERNAL_DOWNLOADS (HIGH): The skill utilizes
npxto dynamically fetch and execute theskillspackage from npm and facilitates downloading additional skills from GitHub repositories. - REMOTE_CODE_EXECUTION (HIGH): The skill instructs the agent to use the
-yflag withnpx skills add, which skips all safety confirmation prompts during remote code installation. This allows for the silent installation and potential execution of malicious scripts from unverified third-party sources. - COMMAND_EXECUTION (MEDIUM): User queries are directly interpolated into shell commands (e.g.,
npx skills find [query]), creating a command injection vulnerability if the agent does not sanitize characters like;,&, or|. - PROMPT_INJECTION (LOW): The skill presents an indirect prompt injection surface. Evidence: 1. Ingestion points: User queries and external search results. 2. Boundary markers: Absent for shell interpolation. 3. Capability inventory: Bash execution and remote package installation. 4. Sanitization: None detected for user or external data.
Recommendations
- AI detected serious security threats
Audit Metadata