course-guide

Pass

Audited by Gen Agent Trust Hub on Nov 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data by fetching curriculum manifests and lesson content from an external repository or local files to guide the agent's routing decisions.
  • Ingestion points: Processes README.md, docs/en.md, and multiple JSON manifests (program.json, tracks/*.json, model-context-protocol.json, etc.) from the rohitg00/ai-engineering-from-scratch repository.
  • Boundary markers: Absent. The instructions do not specify delimiters for the fetched content.
  • Capability inventory: The skill is limited to providing links and recommending the invocation of other specific skills (e.g., learn, learn-mcp) via the host's command syntax. It lacks dangerous capabilities like arbitrary code execution, file writing, or credential access.
  • Sanitization: None detected. The skill relies on the integrity of the curriculum source files.
  • [EXTERNAL_DOWNLOADS]: The skill fetches curriculum data from the author's official GitHub repository at raw.githubusercontent.com/rohitg00/ai-engineering-from-scratch/. This is a vendor-owned resource used for legitimate curriculum synchronization.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 4, 2026, 05:48 AM