#ai-research

Live, open-source benchmark for comparing AI coding agents on real GitHub issues

Open research framework and practical toolkit for studying interaction-shaped behavior, trajectory reconstruction, and long-horizon human–AI collaboration.

Local-first research control plane for gated source packs, claim truth, contradiction handling, and long-running AI synthesis

AI Research Agent 2026 🔍 - Autonomous Deep Web Explorer & Adversarial Synthesis

Agentic AI portfolio: planning agents, multi-agent orchestration, RAG, memory systems, semantic routing, evaluation pipelines, and business workflow automation.

Professional multi-agent benchmark library for evaluating LLMs in 23 strategy games — grid, board, social deduction, cards & game theory

Keep It Real — a schema-level deferral channel that makes LLM brevity safe. Code, data, and pre-registered protocols: a 400-conversation replication sweep, a placebo control, and a blinded human study (n=46).

Evaluator-grounded autonomous prompt evolution. Genetic algorithms + runtime validation + meta-evolution to evolve prompts that generate better AI agent code. 150 generations, 862/1000 best score.