gzh-design-skill-wechat-markdown-formatter
Warn
Audited by Gen Agent Trust Hub on Jul 7, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run local Python scripts (
scripts/component_lint.pyandscripts/validate_gzh_html.py) to validate content and component integrity. These scripts are part of the skill's infrastructure but represent an execution surface for local code. - [EXTERNAL_DOWNLOADS]: The installation and update instructions direct users to fetch resources from a non-whitelisted GitHub repository (
github.com/isjiamu/gzh-design-skill). - [PROMPT_INJECTION]: The skill processes untrusted input data (Markdown, DOCX, and PDF files) to generate formatted output. This represents an indirect prompt injection surface where malicious instructions embedded in source documents could attempt to influence the agent's behavior.
- Ingestion points:
SKILL.md(mentions processingarticle.md,report.docx, and.pdffiles). - Boundary markers: No explicit boundary markers or instructions to ignore embedded commands are specified for the conversion pipeline.
- Capability inventory: The agent has capabilities to execute Python scripts, read various file formats, and write HTML files to the disk.
- Sanitization: While the skill claims to perform 'Dual validation' for platform compliance, it does not specify sanitization or filtering of natural language instructions within the source text.
Audit Metadata