cognithor and cortex
These are competitors—both offer local-first agent operating systems with memory and tool use, but Cognithor provides significantly more integrations (16 LLM providers, 17 channels, 53 MCP tools) while Cortex emphasizes simplicity (pure Python + SQLite, no LangChain), making them alternative choices rather than complementary solutions.
About cognithor
Alex8791-cyber/cognithor
Cognithor - Agent OS: Local-first autonomous agent operating system. 16 LLM providers, 17 channels, 53 MCP tools, 5-tier memory, A2A protocol, knowledge vault, voice, browser automation. Python 3.12+, Apache 2.0.
Implements a hierarchical agent architecture with a 6-phase gateway initialization system and PGE-Trinity processing loop, supporting autonomous task decomposition and multi-agent coordination through an A2A protocol. Provides a Flutter cross-platform command center with GEPA pipeline visualization, resource monitoring dashboards, and voice/chat interfaces alongside 123 MCP-based tools spanning file operations, web scraping, and system management. Emphasizes reproducibility through HMAC/Ed25519-signed audit trails, RFC 3161 timestamping, and GDPR compliance tooling while remaining fully local-first with optional cloud LLM integration.
About cortex
timmeck/cortex
Self-hosted AI Agent OS — Persistent agents with memory, multi-agent coordination, RAG, tool use, and Ollama/Claude support. Pure Python + SQLite, no LangChain.
Scores updated daily from GitHub, PyPI, and npm data. How scores work