Projects
Public software with source code, design decisions and reproducible results. Each project page explains the problem, demonstrated capabilities and limitations.
Citable Publication
Deterministic HTML, Markdown and JSON publishing from one corpus, with integrity verification before release and reproducible faults.
RAG Evaluation Workbench
Reproducible RAG evaluation with BM25, dense and hybrid retrieval, local CPU inference, and separate retrieval and answer metrics.
Agent Replay Runtime
Durable tool execution with event replay, idempotency keys and recovery after real process crashes. Includes effect traces and cases that stop instead of retrying automatically.
Agent Injection Evals
Reproducible prompt-injection evaluation with a pinned local model, simulated tools and separate metrics for task capability, attack attempts and realized effects.
MCP Contract Harness
Read-only MCP server and independent clients for checking stdio negotiation, pagination, errors and cancellation, with actual wire messages and reproducible reports.