genpark-cross-channel-marketing-data-joiner

Warn

Audited by Gen Agent Trust Hub on Aug 1, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill clones source code from a remote repository at https://github.com/alphaparkinc/genpark-cross-channel-marketing-data-joiner-skill.git. This source is not categorized as a trusted vendor.
  • [REMOTE_CODE_EXECUTION]: The installation instructions direct users to install the package directly from a remote Git URL using 'pip install git+https://...'. This method allows for the execution of the package's setup scripts from an unverified external source at installation time.
  • [DATA_EXFILTRATION]: The skill is designed to access and use sensitive API keys and access tokens stored in environment variables (e.g., GOOGLE_ADS_API_KEY, FB_ADS_API_KEY, SHOPIFY_API_KEY, SQUARE_ACCESS_TOKEN). It performs network operations to external service endpoints, including third-party advertising platforms and retail APIs, as well as a custom inventory endpoint (api.warehouse.example.com).
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by ingesting data from external sources.
  • Ingestion points: The skill fetches data from programmatic ad platforms (Google Ads, Facebook Ads, DV360) and retail systems (Shopify, Square) via its connector components.
  • Boundary markers: The skill documentation does not provide boundary markers or instructions to isolate the ingested data from the agent's logic.
  • Capability inventory: The skill possesses network capabilities to communicate with APIs and file-writing capabilities to export CSV and PDF reports.
  • Sanitization: No sanitization or validation of external data is mentioned, which could allow malicious content in marketing data to influence the agent's actions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Aug 1, 2026, 12:27 PM
Security Audit — agent-trust-hub — genpark-cross-channel-marketing-data-joiner