fanqie-webnovel-prose
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill provides a local Python utility (
scripts/check_prose.py) for analyzing prose rhythm and identifying clichés. The script uses standard libraries for local text processing and does not perform network operations or access sensitive system resources. - [INDIRECT_PROMPT_INJECTION]: The skill involves processing user-supplied drafts and outlines. While this is a standard vector for indirect prompt injection, the risk is negligible as the skill lacks high-privilege capabilities and the script logic is limited to non-executable text analysis.
- [METADATA_POISONING]: The
agents/openai.yamlfile contains character encoding errors (mojibake) in the display fields. This appears to be a technical misconfiguration resulting from a character set mismatch (likely GBK interpreted as UTF-8) and does not present a security risk.
Audit Metadata