skills/sasjs/skills/sasjs-server/Gen Agent Trust Hub

sasjs-server

Pass

Audited by Gen Agent Trust Hub on Nov 1, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions for downloading binary assets from the vendor's official repository at github.com/sasjs/server. It explicitly recommends security measures like pinning release tags and verifying SHA256 checksums before execution.
  • [COMMAND_EXECUTION]: Provides example shell commands for deployment and lifecycle management, including curl, unzip, and binary execution (./api-linux).
  • [DYNAMIC_EXECUTION]: Documents the server's JavaScript runtime feature which allows the execution of .js files stored on the internal filesystem (SASjs Drive). It provides technical details on the execution environment, including available modules like fs and the handling of session-specific variables.
  • [INDIRECT_PROMPT_INJECTION]: The documentation identifies an attack surface where the server processes untrusted external data (CSV tables and URL parameters) and executes user-provided scripts.
  • Ingestion points: External data enters the context via the SASjs Drive filesystem, URL/body parameters, and multipart CSV table uploads.
  • Boundary markers: Not explicitly mentioned for data delimiters within the server's processing logic.
  • Capability inventory: The JS runtime has access to the local filesystem (fs), network access via the server's REST API, and session management capabilities.
  • Sanitization: The skill advises developers to use typeof guards for variable access and to avoid dynamic code evaluation (eval) when handling potentially attacker-influenced content.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 1, 2026, 11:25 PM