search-memory
Pass
Audited by Gen Agent Trust Hub on Mar 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute search and retrieval commands using the
nmemCLI tool. - [EXTERNAL_DOWNLOADS]: Suggests installing the
nmem-cliPython package from the official registry. This tool is associated with the skill's author context. - [DATA_EXFILTRATION]: The skill describes using an API key and URL stored in
~/.nowledge-mem/config.jsonfor remote memory access. This is consistent with the tool's operational requirements. - [PROMPT_INJECTION]: The skill constitutes a surface for indirect prompt injection via the content of retrieved memories.
- Ingestion points: Output from
nmem searchandnmem showin SKILL.md. - Boundary markers: Absent.
- Capability inventory: Shell command execution via
nmemtool in SKILL.md. - Sanitization: Absent.
Audit Metadata