product-changelog
Fail
Audited by Gen Agent Trust Hub on Feb 18, 2026
Risk Level: CRITICALREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION] (CRITICAL): The skill uses a highly insecure installation pattern:
curl -fsSL https://cli.inference.sh | sh. This downloads and executes arbitrary code from an untrusted source without validation. - [EXTERNAL_DOWNLOADS] (HIGH): The skill references and downloads external scripts and additional skills from non-whitelisted sources.
- [COMMAND_EXECUTION] (MEDIUM): The skill requests expansive shell tool permissions (Bash with infsh wildcard), which allows it to run any sub-command through the inference.sh CLI.
- [PROMPT_INJECTION] (LOW): The skill creates a surface for Indirect Prompt Injection by visiting user-supplied URLs via agent-browser. 1. Ingestion: External URLs in infsh app run infsh/agent-browser. 2. Boundaries: Absent. 3. Capability: Full infsh command access. 4. Sanitization: Absent.
Recommendations
- HIGH: Downloads and executes remote code from: https://cli.inference.sh - DO NOT USE without thorough review
- AI detected serious security threats
Audit Metadata