Skills
Navigate
skills/jeremylongshore/tons-of-skills-marketplace/brightdata-deploy-integration/Gen Agent Trust Hub

brightdata-deploy-integration

Pass

Audited by Gen Agent Trust Hub on Sep 13, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The serverless function template in api/scrape.ts contains a vulnerability surface for Server-Side Request Forgery (SSRF). 1. Ingestion points: The url parameter is accepted directly from req.body.url in SKILL.md. 2. Boundary markers: Absent; no delimiting or instructions to ignore embedded content are provided. 3. Capability inventory: The function utilizes the axios.get method to perform network requests in SKILL.md. 4. Sanitization: Absent; the provided input is used to make a network request without any validation or allow-listing.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: Code examples provided for deployment in api/scrape.ts and api/health.ts intentionally disable SSL certificate verification. 1. Evidence: The snippets use httpsAgent: new https.Agent({ rejectUnauthorized: false }), which bypasses standard certificate validation and increases the risk of Man-in-the-Middle (MitM) attacks during proxy communication.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 13, 2026, 01:41 AM
Security Audit — agent-trust-hub — brightdata-deploy-integration