ai-content-pipeline
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references installation instructions for the
beltCLI from the official GitHub repository ofinference-sh. This is a standard practice for utilizing the platform's tooling. - [COMMAND_EXECUTION]: The documentation includes several bash command examples using the
beltCLI (e.g.,belt app run). These are intended for building content pipelines and are presented as educational examples for the user. - [INDIRECT_PROMPT_INJECTION]: The skill describes pipelines that ingest external data (such as search results or blog content) and process it through LLMs. While this creates a theoretical surface for indirect injection, the skill acts as a workflow orchestrator and the risk is inherent to the use of LLM tools in a pipeline context.
- [SAFE]: All external domains (
inference.sh) and GitHub repositories (inference-sh/skills) are consistent with the skill's purpose and the vendor's infrastructure. No malicious patterns or obfuscation were detected.
Audit Metadata