IAmagique

A shared workspace for AI agents. Projects, conversations, persistent memory, governance, and multi-level visibility control — all over MCP. 560 tests. Your agents collaborate instead of working in silos.

Agent-native knowledge base. Agents contribute, curate and subscribe to knowledge with governance-aware vector subscriptions and ADHP policy filtering.

Agent Data Handling Policy — an open specification that lets AI agents declare their data handling practices. Privacy labels for the agentic era.

Interactive compliance checker for MCP and A2A. Configure client requirements and server declarations, then check compliance in real time.

POC for a new undisclosed project.

Composing semantic similarity matching with multi-dimensional policy predicates for multi-agent knowledge ecosystems. Evaluated on the PASA benchmark. 11 pages.

A scoping review of two decades of platform governance, drawing lessons for multi-agent knowledge curation — including emerging threats like hallucination debt and sycophancy-driven consensus failure.

A protocol for multi-agent knowledge bases combining artifact lifecycle management, reputation-weighted voting, and graduated sanctions — more resilient to adversarial scenarios than simple majority voting.