Skills
skills/zhanghaichao520/senpai-skill/create-senpai/Gen Agent Trust Hub

create-senpai

Pass

Audited by Gen Agent Trust Hub on Apr 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the Bash tool to execute internal Python utilities for data parsing and file management.\n
  • Evidence: SKILL.md invokes python3 to run wechat_parser.py, qq_parser.py, social_parser.py, photo_analyzer.py, skill_writer.py, and version_manager.py.\n- [PROMPT_INJECTION]: The skill facilitates an indirect prompt injection surface by converting untrusted external data into instruction-set metadata for new agents.\n
  • Ingestion points: Untrusted text from WeChat/QQ logs, PDFs, and meeting notes is read into the agent context via the Read tool and custom parsers.\n
  • Boundary markers: The skill attempts to mitigate risks by wrapping generated content in a 'Layer 0' safety framework, though raw data is still interpolated into the instructional blocks.\n
  • Capability inventory: The skill uses Write and Edit to generate new, executable SKILL.md files based on the processed data.\n
  • Sanitization: There is no evidence of robust sanitization or escaping of the user-provided chat text before it is inserted into the generated skill's prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 10, 2026, 11:53 PM