picgo-upload
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to interact with the PicGo ecosystem, including
node scripts/gui-upload.mjsfor the desktop application,picgo uploadfor the CLI, andnpm install picgo -gfor setup. These commands are the primary method of providing the skill's functionality. - [EXTERNAL_DOWNLOADS]: The skill guides the user to install the
picgopackage from the official NPM registry if it is missing. This refers to a well-known service and is a standard requirement for the tool's operation. - [INDIRECT_PROMPT_INJECTION]: The skill handles user-provided files (images, PDFs, Zips) which could theoretically contain malicious instructions designed to influence the agent during metadata extraction or description generation. The skill mitigates this by providing clear operational boundaries and requiring user confirmation for sensitive content.
- [PRIVILEGE_ESCALATION]: While the skill mentions
sudo npm installas a solution for permission errors, it explicitly commands the agent NOT to run this automatically. It instead requires the agent to present the command for the user to execute manually, ensuring the user maintains control over elevated permissions.
Audit Metadata