twitter-automation
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references the installation of the
beltCLI using thenpx skills add belt-sh/clicommand and links to installation instructions hosted on a public GitHub repository belonging to theinference-shorganization. These resources are standard requirements for the tool's stated purpose. - [COMMAND_EXECUTION]: The skill defines a set of commands for the
beltCLI to interact with the Twitter API, includingbelt loginfor authentication andbelt app runfor executing specific social media tasks. The frontmatter includes anallowed-toolsconstraint (Bash(belt *)) which restricts the agent's execution environment to only allow thebeltcommand, serving as a security control. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided strings for tweet text, image prompts, and video descriptions. This creates a surface where external content is interpolated into API calls, which is a standard feature for social media automation tools. The documentation includes examples of handling JSON input files for structured data management.
- [DATA_EXPOSURE]: The skill uses a
belt loginflow to manage credentials. It does not contain hardcoded API keys, tokens, or sensitive local file path access. It guides users to store configuration ininput.jsonfiles for batch processing, which is consistent with standard development practices.
Audit Metadata