llmem-setup
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches installation scripts and software from external sources.
- Fetches a setup script from the author's GitHub repository (
MichielDean/LLMem). - Fetches the Ollama installation script from the
ollama.comdomain. - [REMOTE_CODE_EXECUTION]: The skill executes downloaded scripts directly within the shell environment.
- The
setup.shscript from the project repository is piped tobashfor installation. - The official Ollama installation script is piped to
shto set up the local embedding provider. - [INDIRECT_PROMPT_INJECTION]: The skill manages a memory system that retrieves stored data into the agent's context, which could potentially contain untrusted instructions.
- Ingestion points: Data enters the context via the
llmem searchcommand output. - Boundary markers: Not explicitly mentioned; memories are injected as raw text into the session.
- Capability inventory: The agent has the ability to execute shell commands, install packages via npm/pip, and configure systemd services.
- Sanitization: There are no instructions regarding the sanitization or filtering of retrieved memory content.
Audit Metadata