brand-voice
Pass
Audited by Gen Agent Trust Hub on Nov 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run a local Python script (
python3 docs/brand/lint.py) and various vendor-owned utility tools vianpx(e.g.,sshlg-skills,sheleg-design-skill,task-pipeline). These tools are used for project-specific linting, brand initialization, and text humanization passes. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external project files, such as UI source code and marketing content, as defined in the project's
README.md. This ingestion process creates a potential attack surface for indirect prompt injection if the analyzed files contain malicious instructions aimed at the agent's behavior during the linting workflow. - Ingestion points: Source files defined by globs in
README.md(e.g.,src/**/*.{ts,tsx},content/**/*.{md,mdx}). - Boundary markers: The skill does not define explicit delimiters to separate analyzed text from agent instructions.
- Capability inventory: The skill has the ability to execute shell commands (
python3,npx) and read/write project files. - Sanitization: No explicit sanitization of ingested content is described before processing.
- [EXTERNAL_DOWNLOADS]: The documentation references and links to several external resources for marketing research and algorithmic analysis, such as the
Jakeschincariol/instagram-agent-skillGitHub repository and official documentation for search engine crawlers. These references are used solely for informational and logic guidance and do not involve unauthorized remote code execution.
Audit Metadata