What problem does it solve? Physics derivations often contain hidden assumptions, unjustified approximations, and silent convention errors (sign flips, missing factors of 2pi) that propagate undetected. This Skill enforces a rigorous derivation workflow where every step is verified, every assumption is explicit, and every approximation is bounded with error estimates. ## Core Features & Use Cases - Convention Locking: Loads project convention locks (metric signature, Fourier convention, normalization) and asserts them per-step via machine-readable ASSERT_CONVENTION markers to catch convention drift. - Systematic Verification: Performs dimensional analysis, limiting-case checks, symmetry preservation tests, and numerical spot-checks at each major derivation step. - Documented Output: Produces a complete, self-contained derivation document with numbered assumptions, boxed final result, error analysis, and connection to known results, saved to phase-scoped or standalone analysis directories. - Use Case: Derive the effective mass from a self-energy expression in a quantum field theory project, with each intermediate result checked against the free-theory limit and Peskin & Schroeder reference values. ## Quick Start Ask the AI to derive an equation, for example: derive the dispersion relation for the current system with full step-by-step verification.
