learn
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill fetches course text (markdown) and quiz data (JSON) from an external GitHub repository, which could potentially contain instructions designed to manipulate the agent's behavior.
- Ingestion points: Fetches content from
https://raw.githubusercontent.com/fancyboi999/ai-engineering-from-scratch-zh/main/<path>defined inSKILL.md. - Boundary markers: No specific delimiters or instructions are provided to the agent to treat the fetched content strictly as data or to ignore any embedded commands.
- Capability inventory: The agent can write to local progress logs (
LEARNING.md) and is instructed to suggest host-specific command invocations (e.g.,/skill-name). - Sanitization: There is no evidence of content sanitization or validation of the fetched markdown/JSON before it is processed and presented to the learner.
- [EXTERNAL_DOWNLOADS]: The skill is designed to download pedagogical materials from the author's public repository (
fancyboi999/ai-engineering-from-scratch-zh) when local files are unavailable. This includes markdown documentation and structured quiz data used to drive the interactive learning session.
Audit Metadata