deltasci

The grounding layer for AI-assisted research: scan the real prior art around an idea, gauge how crowded or open the gap is, and verify every citation against the source of truth. You (the agent) supply the discovery intelligence — writing search queries, judging relevance, reasoning about the gap — while the deterministic deltasci engine fetches real records and runs the citation checks. No LLM ever sits in the trust path: a citation is "verified" only when the engine says so, never from your memory. Use when a researcher wants to ground an idea, a related-work section, or a whole paper against the real literature — find the closest existing work, see where the genuine opening is, and catch fabricated / wrong-paper / unsupported citations.

Two-perspective co-reasoning for AI4Science hypothesis generation. Runs a structured 4-round dialogue between a domain scientist (parameterized by a domain pack) and an ML engineer, producing a grounded, falsifiable research hypothesis that is honest about the AI's training-distribution edges. Domain-agnostic via pluggable packs (biomed, materials, climate, or your own). Use when a researcher has a vague idea and wants to turn it into a defensible, evaluable hypothesis with explicit handoffs for the things only the researcher can know.