What problem does it solve? Writing and optimizing solutions for ICFPC 2026 littleman problems requires a verified local workflow: guessing instruction semantics or checking behavior in the production editor leads to wrong submissions. This Skill enforces a disciplined loop of reading problem specs, writing .man programs, validating them locally with the lman simulator, optimizing footprint and tick scores, and submitting through the team proxy. ## Core Features & Use Cases - Local verification with lman: Runs structural checks, per-tick traces, and full public test suites against the official littleman interpreter wrapper, with coordinate-annotated errors. - Score optimization guidance: Prioritizes footprint reduction over tick reduction for footprint-tick scoring, with BFS-based instruction sequence search and grid-size analysis. - Per-problem organization: Keeps each problem in problems/<slug>/ with named .man variants and a NOTES.md documenting the program, score breakdown, and register transition tables. - Use Case: Ask to solve the triangle problem; the Skill reads the spec via lman, writes problems/triangle/triangle-9x9.man, runs lman test until all public cases pass, checks rank against official standings, and submits via the team proxy. ## Quick Start Ask the assistant to solve the ICFPC littleman problem 'triangle' and verify it locally with lman before submitting through the team proxy.
