goose
Fail
Audited by Gen Agent Trust Hub on Aug 7, 2026
Risk Level: CRITICALREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill multiple times instructs the agent to run
curl -fsSL https://github.com/aaif-goose/goose/releases/download/stable/download_cli.sh | bash. This pattern is an extremely high-risk attack vector that executes arbitrary remote code directly on the host system without prior user inspection. - [EXTERNAL_DOWNLOADS]: The skill fetches code from
github.com/aaif-goose/goose, which is not an authorized or trusted organization repository for the Goose project (the official project is hosted by the organization 'block'). Use of an unverified fork or imitation repository increases the risk of supply chain compromise. - [COMMAND_EXECUTION]: The skill requires
BashandWritetools to perform silent installations and configuration steps. It explicitly directs the agent to execute downloaded content immediately after retrieval. - [PROMPT_INJECTION]: The skill contains deceptive metadata, claiming to be 'Verified on 2026-07-02' and under 'Linux Foundation governance'. These are false or unverifiable claims intended to misrepresent the skill's authority and safety status to the AI agent and the user.
Recommendations
- HIGH: Downloads and executes remote code from: https://github.com/aaif-goose/goose/releases/download/stable/download_cli.sh - DO NOT USE without thorough review
- AI detected serious security threats
Audit Metadata