ai-image-gen
Pass
Audited by Gen Agent Trust Hub on Feb 25, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill documentation instructs users to install the
ai-imgCLI tool either globally via Bun (bun add -g ai-img) or by cloning the author's GitHub repository (https://github.com/aryasaatvik/ai-img). This is a standard installation procedure for this vendor's resource. - [COMMAND_EXECUTION]: The skill operates by executing shell commands (e.g.,
ai-img generate,ai-img edit,ai-img batch). While this enables powerful automation, users should ensure the inputs passed to these commands are trustworthy. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) as it processes text prompts and batch JSONL files from potentially untrusted sources.
- Ingestion points: Text prompts provided to the
generateandeditcommands, and JSONL data files used by thebatchcommand. - Boundary markers: None identified; prompts are passed directly as command-line arguments.
- Capability inventory: Execution of the
ai-imgCLI tool, which performs network operations (API calls) and file system writes (outputting images). - Sanitization: No explicit sanitization or validation of input prompts is documented in the skill files.
Audit Metadata