Skills
skills/sickn33/antigravity-awesome-skills/langchain-architecture/Gen Agent Trust Hub

langchain-architecture

Pass

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill documents architectural patterns for RAG and AI agents that process external data, creating a surface for indirect prompt injection.\n
  • Ingestion points: Uses TextLoader and DirectoryLoader to ingest files from documents.txt and ./docs directories in SKILL.md.\n
  • Boundary markers: Provided patterns lack delimiters or instruction isolation techniques.\n
  • Capability inventory: Includes capabilities for web searching (serpapi), math execution (llm-math), database access, and email dispatch.\n
  • Sanitization: No data validation or sanitization routines are provided in the examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 28, 2026, 09:04 AM