"author:SkilledDragonCorp" matching MCP servers:

Scores proposed tool calls with a decision model before execution, returning allow, ask, or block decisions with safety, production, and prompt-injection probabilities so agents can gate risky actions and fail closed on scorer errors.

Enables agents to add project notes as sources, list them, and ask questions that come back as text with [n] markers plus numbered quotes from the source. Backends are pluggable (local BM25 by default, Gemini File Search, or notebooklm-py) behind a risk gate that filters secrets and personal data before any call leaves the machine.