dbs-wechat-html
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes standard command-line tools
xmllintandrg(ripgrep) for the sole purpose of validating the generated HTML output and ensuring it complies with security and compatibility constraints.\n- [INDIRECT_PROMPT_INJECTION]: While the skill processes user-provided Markdown (untrusted data), it implements several defensive measures:\n - Ingestion points: Processes Markdown text provided by the user or read from local files (defined in SKILL.md).\n
- Boundary markers: Explicitly instructs the agent to act as a layout engine only and not to modify or polish the content, creating a functional boundary.\n
- Capability inventory: Limited to file system writes in a specific output directory and the execution of validation commands (SKILL.md).\n
- Sanitization: Includes mandatory instructions for HTML special character escaping and a secondary security scan to remove scripts, styles, and external URLs (SKILL.md).\n- [SAFE]: The skill explicitly forbids the inclusion of JavaScript, external CSS, or remote resources in its final output, which significantly reduces the risk of cross-site scripting (XSS) or data exfiltration via the generated HTML.
Audit Metadata