film-prop-sheet
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill defines a workflow that involves executing local shell scripts (
batch.sh,poll.sh) and a CLI tool (cohub) to handle image generation and asset polling. These scripts are expected to be located in a local directory defined by the agent at runtime. - [INDIRECT_PROMPT_INJECTION]: The skill provides templates for shell commands that incorporate user-controlled data such as
<slug>and prop descriptions. While this is a common pattern for such workflows, it represents a potential injection surface if inputs are not validated. - Ingestion points: User-provided strings for placeholders like
<slug>,<prop>,<material>, and<team>inSKILL.md. - Boundary markers: None; the placeholders are directly interpolated into the command strings.
- Capability inventory: Execution of subprocesses via bash scripts and the
cohubcommand. - Sanitization: None; the instructions do not include specific validation steps for the input placeholders.
Audit Metadata