add-remote-skill
Pass
Audited by Gen Agent Trust Hub on Feb 28, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Clones skill directories from user-provided GitHub URLs using sparse checkout. This facilitates the integration of remote repository content into the local marketplace environment.
- [COMMAND_EXECUTION]: Executes a local TypeScript script
bin/add-remote-skill.tsusing thenpx tsxcommand to handle URL parsing, directory cloning, and file system operations. - [PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection by ingesting data from external repositories.
- Ingestion points: Content from remote GitHub repositories specified by user URLs.
- Boundary markers: None identified; the skill clones and integrates external
SKILL.mdfiles directly. - Capability inventory: The skill can perform filesystem operations, modify file headers, and execute local scripts via
npx. - Sanitization: Validates GitHub URL formats but does not mention sanitizing or filtering the content of the cloned markdown files.
Audit Metadata