chart
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [Data Exposure & Exfiltration] (LOW): The skill transmits user-provided data to
https://quickchart.iofor chart generation. This domain is not on the trusted whitelist. While appropriate for the skill's function, it exposes data to a third-party service. - [Indirect Prompt Injection] (LOW): The skill allows the agent to read local files based on user-provided paths and include that data in the external API call. This creates a surface for reading and exfiltrating sensitive local files if the agent is not specifically instructed to validate paths or data content. (Evidence: Ingestion point in file-based data input; Capability inventory includes Read and Bash(curl); Boundary markers and sanitization are absent).
- [Command Execution] (LOW): The skill's instructions involve constructing a
bashcommand (curl) that includes user-provided strings (labels, titles). This presents a risk of command injection or argument manipulation if the agent does not correctly escape the input when formatting the JSON payload for the shell.
Audit Metadata