learn-reverse-engineering
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches curriculum configuration, documentation, and exercise code from the vendor's GitHub repository at https://raw.githubusercontent.com/fancyboi999/ai-engineering-from-scratch-zh/.
- [COMMAND_EXECUTION]: The instructions require the agent to run local commands such as 'python3 --version', 'clang', and 'llvm-objdump' to verify the environment and analyze user-provided binaries.
- [REMOTE_CODE_EXECUTION]: The skill specifies reading and executing code from the 'code/' directory of the remote repository to perform experiments during the lessons.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data including 'quiz.json', 'docs/zh.md', and user-compiled binaries. This ingestion surface is a vector for indirect instructions. 1. Ingestion points: quiz.json and docs/zh.md (SKILL.md). 2. Boundary markers: No delimiters or warnings present. 3. Capability inventory: Execution of python3 and llvm-objdump (SKILL.md). 4. Sanitization: None identified.
Audit Metadata