The Agent Skills Directory

[SAFE]: The skill serves as a guide for integrating a semantic memory service via the MCP protocol. It contains no instructions for overriding agent behavior or bypassing safety filters.
[DATA_EXFILTRATION]: All network activity is directed toward a user-defined local API endpoint (defaulting to localhost:8000). No sensitive credentials or private files are accessed or transmitted to external domains.
[REMOTE_CODE_EXECUTION]: The skill documents the local execution of a Python-based MCP server and the use of Docker for hosting the memory service. These are legitimate setup procedures and do not involve downloading or executing code from untrusted remote sources.

mesh-memory