TGMS blog

Engineering notes from building an agent-native bi-temporal graph store. Every post ships with runnable artifacts and determinism receipts — no result without a reproduction path.
#1

Why your agent needs two kinds of time

Correction probes: the question class that snapshot and RAG systems cannot express — and what happened when we measured it. Baselines score zero at every model size; the bi-temporal system doesn’t care how big the model is, because the model isn’t doing the temporal work.

bi-temporalbenchmarks
July 2026 · published
#2

Statically valid, semantically doomed

A 7B model met our operator schemas: perfectly valid plans, zero successful executions. Models invent output paths no input schema constrains. On output contracts — the one-line class of check that took execution success from 0.00 to 1.00 — and why tool manuals must be checkable, not just readable.

small modelstool contracts
coming soon
#3

Your verifier believes correct arithmetic on incomplete evidence

A live demo caught our claim verifier approving a count computed over a truncated result page — the value matched, the evidence didn’t. On truncation taint, transitively poisoned steps, and treating evidence quality as a lattice instead of a boolean.

verificationwar story
coming soon