ai-content-pipeline
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
infshCLI tool through Bash to interact with various AI models and media processing services. - [EXTERNAL_DOWNLOADS]: It references several related skills within the
inference-shorganization (e.g.,inference-sh/skills@agent-tools) to be installed vianpx. These are vendor-provided resources associated with the skill author. - [PROMPT_INJECTION]: The skill documents multi-step pipelines where the output from one model (such as a generated script or summary) is used as the input for another model (such as text-to-speech or image generation).
- Ingestion points: Untrusted data enters the context through model-generated scripts and user-provided prompts within
SKILL.mdexamples. - Boundary markers: None. The examples do not use delimiters or instructions to ignore embedded commands when passing text between models.
- Capability inventory: The skill uses the Bash tool to execute
infshcommands (SKILL.md). - Sanitization: None. The bash examples show direct interpolation of variables into command strings without escaping or validation.
Audit Metadata