javascript-sdk
Fail
Audited by Gen Agent Trust Hub on Apr 14, 2026
Risk Level: HIGHREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: Documentation examples in 'references/tool-builder.md' and 'references/agent-patterns.md' demonstrate using the 'eval()' function to execute code strings provided directly by the AI agent as tool arguments. This is a critical vulnerability that allows for arbitrary code execution if the agent is manipulated through prompt injection.
- [PROMPT_INJECTION]: The skill facilitates the creation of agents that ingest untrusted data from external sources, presenting a surface for indirect prompt injection. (1) Ingestion points: 'internalTools().webSearch(true)' and 'client.uploadFile' described in 'SKILL.md'. (2) Boundary markers: No delimiters or isolation techniques are demonstrated in the provided patterns to separate data from instructions. (3) Capability inventory: The skill supports high-impact capabilities including built-in code execution and custom webhook tools. (4) Sanitization: No sanitization or validation logic is provided for processing tool arguments derived from agent-controlled context.
Recommendations
- AI detected serious security threats
Audit Metadata