ai-fusion-daily
Pass
Audited by Gen Agent Trust Hub on Jun 23, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTIONREMOTE_CODE_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes untrusted content from external news sources (AI HOT API and GitHub Trending), creating an indirect prompt injection surface.
- Ingestion points: External news content and GitHub project descriptions.
- Boundary markers: Uses specific HTML comments (e.g., EMBEDDED_DATA_START) to delimit managed data sections in index.html.
- Capability inventory: The skill executes node build.js for data processing and performs local file writes.
- Sanitization: The indexing script includes a text stripping function that removes HTML tags from fetched content before it is added to the dashboard.
- [COMMAND_EXECUTION]: The skill requires running node build.js to aggregate report data and update the central index. This represents a local script execution as part of the tool's core functionality.
- [EXTERNAL_DOWNLOADS]: Fetches data from GitHub, which is recognized as a well-known and established service.
- [REMOTE_CODE_EXECUTION]: A sample report file (20260623.html) contains a text description of a curl | bash installation command for a documented AI tool. This is present as informative documentation in the report and is not an instruction for the agent to execute code.
Audit Metadata