fin-opencli-reader

Warn

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [UNVERIFIABLE_DEPENDENCIES]: The skill requires installing @jackwener/opencli via npm and uses npx super-agent and npx skills, which are third-party tools from unverified sources.
  • [COMMAND_EXECUTION]: The skill requires the execution of shell commands for setup and data retrieval, which can be misused if the tools are compromised.
  • [PRIVILEGE_ESCALATION]: Using npm install -g often necessitates elevated privileges (e.g., sudo), increasing the potential impact of the external installation.
  • [INDIRECT_PROMPT_INJECTION]: The skill summarizes data from untrusted websites like Reddit and HackerNews.
  • Ingestion points: Output from the opencli command when fetching content from external websites.
  • Boundary markers: No delimiters or instructions to ignore embedded commands are present.
  • Capability inventory: Shell command execution via npm and opencli.
  • Sanitization: No evidence of filtering or sanitization of retrieved web data.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Aug 25, 2026, 09:39 AM