bootloops-ai/bootloops-protocols

The standard a computed result must meet before it may be called done. Use whenever an agent claims an integral is solved, a fit has closed, a formula is identified, or any quantitative answer is ready — the claim is not done until it passes this gate.

Discipline for integer-relation searches (PSLQ, LLL) that turn high-precision digits into exact constants. Use whenever recognizing a numerical value as a closed form, fitting boundary constants, or reporting that no closed form exists.

Provenance rules that keep checks independent. Use when setting up any comparison, oracle, or reference value — and whenever tempted to certify a result with machinery that helped produce it.

Synthetic-truth controls for analysis pipelines. Use before running any statistical fit, solver, audit, or search on real data — the pipeline must first prove it can recover a known planted answer and catch a deliberately corrupted input.

Sources-only discipline for any instrument or agent that reads documents and asserts facts from them — literature audits, record replays, data extraction, citation of prior work.

Compute-planning rules. Use before launching any calculation whose runtime is not already measured — and whenever an agent proposes a long run, an overnight grind, or a bigger machine as the way forward.

How the toolkit is consulted, extended, and documented. Use before writing any new code — and after, when a genuinely new instrument has earned its place.