memory-config
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches package version metadata from the official PyPI API and plugin configuration manifests from the vendor's official GitHub repository. These operations target well-known and trusted sources.
- [REMOTE_CODE_EXECUTION]: Employs a pipe from curl to python3 to parse version strings from a JSON response. This is a false positive for malicious remote code execution, as the Python code is a static inline script used as a utility for data parsing, and the data source is a trusted package registry.
- [COMMAND_EXECUTION]: Utilizes standard system commands including git, uv, and the vendor's own memsearch CLI to perform health checks and update configuration settings.
- [INDIRECT_PROMPT_INJECTION]: The skill processes memory journal files which contain historical conversation data, creating a potential surface for indirect injection.
- Ingestion points: Reads markdown files from the .memsearch/memory/ directory.
- Boundary markers: No explicit delimiters or ignore-instructions warnings are defined in the instruction set.
- Capability inventory: Can modify plugin configuration via memsearch config set and install/update tools via uv.
- Sanitization: Not explicitly implemented; the skill relies on the agent's core safety protocols and its specific focus on MemSearch configuration keys.
Audit Metadata