Skills
skills/yigitkonur/skills-by-yigitkonur/build-langchain-ts-app/Gen Agent Trust Hub

build-langchain-ts-app

Pass

Audited by Gen Agent Trust Hub on May 10, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: A technical guide in references/knowledge-agents.md contains an example that fetches a sample SQLite database from storage.googleapis.com. This is a well-known service (Google Cloud Storage) and the download provides a standard dataset for tutorial purposes.
  • [PROMPT_INJECTION]: Deterministic detectors flagged instruction overrides in references/mcp.md. Analysis confirms these are false positives originating from technical documentation of Model Context Protocol configuration options, such as the onConnectionError: "ignore" parameter.
  • [SAFE]: The skill is a documentation-only resource and does not include executable scripts in its root directory. It explicitly teaches and provides code for security guardrails, including regex-based PII filtering and SQL query validation.
Audit Metadata
Risk Level
SAFE
Analyzed
May 10, 2026, 03:49 PM