product-changelog
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references external installation instructions and dependencies necessary for its functionality.
- It points to a remote markdown file for CLI installation:
https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md. - It suggests the installation of additional tools and skills using the
npx skills addcommand, targetingbelt-sh/cliandinference-sh/skills. - [COMMAND_EXECUTION]: The skill utilizes the
beltCLI tool (authorized in the YAML frontmatter) to perform various tasks. - It includes examples for user authentication (
belt login) and running specific remote applications (belt app run) for image generation, browser automation, and image processing. - [INDIRECT_PROMPT_INJECTION]: The skill serves as a framework for processing external data (product updates and release notes) which constitutes a potential injection surface.
- Ingestion points: The agent processes user-supplied feature descriptions and changelog entries to format them according to the provided templates.
- Boundary markers: No explicit delimiters or instructions to ignore embedded commands are included in the markdown templates.
- Capability inventory: The skill utilizes the
belttool for network-based application execution and visual generation. - Sanitization: No specific filtering or validation of the input content is documented.
Audit Metadata