Tools from TuringSpark to help you build, debug, and ship AI-powered applications faster.
A zero-config CLI proxy that logs every JSON-RPC message between your MCP client and server — with full payloads, timing, and colour-coded direction arrows. Drop it in front of any MCP server in seconds.
npx @turingspark/mcp-debug --stdio -- node server.jsA CLI debugging and testing toolkit for AI agent developers who build on MCP. Record every tool call, inspect traces, diff runs, replay deterministically, and assert in CI — no code changes required.
npx @turingspark/mcp-tracer record --stdio "your-mcp-server"A CLI-first data curation engine for ML engineers. Curates fine-tuning datasets by removing near-duplicates, scoring educational value, and detecting contradictions — reduce dataset size by 50–80% without sacrificing accuracy.
pip install truvaCollect human preference feedback on LLM outputs directly from your Python code. Submit pairs of model responses, pick the better one in a local browser UI, and export a DPO-ready JSONL dataset. Local-first, zero config.
pip install feedloopAn intelligent tutoring extension for OpenClaw that remembers each student across sessions, tracks mastery concept-by-concept, detects prerequisite gaps, and adapts its teaching strategy using spaced repetition.
npx clawtutor setupMore tools coming soon
We're working on more tools for AI developers.