mem0-vercel-ai-sdk
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions to install the
@mem0/vercel-ai-providerpackage and various Vercel AI SDK provider packages from the official npm registry. - [INDIRECT_PROMPT_INJECTION]: The skill's primary functionality involves augmenting LLM prompts with external data retrieved from the Mem0 Platform API. This introduces a surface for indirect prompt injection where the stored memory content could potentially contain instructions intended to influence the model's behavior.
- Ingestion points: Data is ingested into the prompt via memory search results fetched from
https://api.mem0.ai, as detailed inSKILL.mdand the reference documentation. - Boundary markers: The skill documentation shows the use of a system preamble to delineate memories: "These are the memories I have stored. Give more weightage to the question by users... If the memories are irrelevant you can ignore them."
- Capability inventory: The integration code performs network requests to the Mem0 API for searching (
POST /v3/memories/search/) and storing (POST /v3/memories/add/) conversation history. - Sanitization: The provided documentation and code examples do not describe explicit sanitization or filtering of the retrieved memory strings before they are injected into the LLM system prompt.
Audit Metadata