skills/mem0ai/mem0/mem0-vercel-ai-sdk/Gen Agent Trust Hub

mem0-vercel-ai-sdk

Pass

Audited by Gen Agent Trust Hub on Apr 27, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of standard Node.js packages such as @mem0/vercel-ai-provider and ai via npm from the official registry.
  • [COMMAND_EXECUTION]: Instructions guide the user to set environment variables for API keys and perform package installations using standard terminal commands, which is consistent with developer tools.
  • [DATA_EXFILTRATION]: The skill is designed to communicate with the official Mem0 API at api.mem0.ai to store and retrieve conversation memories. This behavior is the primary intended function of the skill and targets a known vendor service.
  • [PROMPT_INJECTION]: The skill implements a pattern for injecting external data (memories) into the system prompt. It employs boundary markers (preamble instructions) to distinguish injected context from user input, mitigating the risk of the agent misinterpreting the retrieved data as direct commands. Ingestion points: retrieveMemories and getMemories fetch data from api.mem0.ai. Boundary markers: 'System Message' preamble with explicit model guidance. Capability inventory: Modification of LLM prompts via the Vercel AI SDK. Sanitization: No client-side filtering of memory content before injection is specified, relying on the platform's handling of prompt context.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 27, 2026, 05:58 AM