brightdata-reference-architecture

Pass

Audited by Gen Agent Trust Hub on Sep 13, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The architecture facilitates the ingestion and processing of untrusted data from external URLs, creating a vulnerability surface for indirect prompt injection.
  • Ingestion points: Untrusted content enters the agent context through the scrape and scrapeWithBrowser methods within src/brightdata/client.ts.
  • Boundary markers: The implementation lacks explicit delimiters or instructions to ignore potential commands embedded within the scraped HTML content.
  • Capability inventory: The skill possesses network capabilities (proxy and API access) and automated execution via node-cron in src/pipeline/scheduler.ts, which could be targeted by an injection attack.
  • Sanitization: While the prose documentation mentions "schema/data-class checks," the provided code snippets do not include specific sanitization or filtering to neutralize malicious instructions embedded in the raw HTML data.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 13, 2026, 01:41 AM
Security Audit — agent-trust-hub — brightdata-reference-architecture