"author:Icaro0310" matching MCP servers:

Exposes local MCP tools for typed, calibrated LLM decisions — gating risky actions, yes/no judging, classification, ordinal rating, and multi-question decisions in one pass. Runs on the model Devin's CLI already provides via ACP (or a fully offline local backend), so agents get calibrated confidence with no extra API key, no Ollama, and no waitlist.

Enables agents and MCP clients to retain, recall, screen, quarantine, approve, supersede, and verify durable memories with session-level provenance. Every write passes a poisoning gate, so suspicious or secret-bearing entries are held for human review rather than silently trusted in later sessions.