skills/mem0ai/mem0/mem0-onboard/Gen Agent Trust Hub

mem0-onboard

Pass

Audited by Gen Agent Trust Hub on May 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local script scripts/setup_coding_categories.py using the system Python interpreter to configure the environment. It also provides a command to initialize the agent using the mem0-cli tool.
  • [EXTERNAL_DOWNLOADS]: The wizard recommends installing the mem0-cli package from PyPI. This is a legitimate utility provided by the vendor (mem0ai) to facilitate API key management.
  • [DATA_EXFILTRATION]: The skill reads project configuration files such as CLAUDE.md, .cursorrules, and AGENTS.md and transmits their content to the Mem0 service via the add_memory tool. This behavior is the primary intended function of the onboarding process to provide the agent with project context.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests content from external files (CLAUDE.md, etc.) which could contain malicious instructions. However, the skill uses the infer=False parameter when calling add_memory, which instructs the service not to interpret the content as new instructions during ingestion, mitigating the risk.
Audit Metadata
Risk Level
SAFE
Analyzed
May 20, 2026, 08:32 PM
Security Audit — agent-trust-hub — mem0-onboard