Guidance for implementing minimal GPT-2 inference in constrained environments (code golf challenges). This skill should be used when implementing neural network inference from scratch, parsing binary checkpoint formats, implementing BPE tokenization, or working on code golf challenges involving ML models.
Install
npx skills add benchflow-ai/skillsbench --skill gpt2-codegolfnpx skills add benchflow-ai/skillsbench --skill gpt2-codegolf -a claude-code -gnpx skills add benchflow-ai/skillsbench --skill gpt2-codegolf -a codex -ggemini skills install https://github.com/benchflow-ai/skillsbench.git --path registry/terminal_bench_2.0/letta_skills_batch/terminal_bench_2_0_gpt2-codegolf/environment/skills/gpt2-codegolfgh skill install benchflow-ai/skillsbench gpt2-codegolfgit clone --depth 1 https://github.com/benchflow-ai/skillsbench /tmp/benchflow-ai__skillsbench
mkdir -p ~/.claude/skills ~/.agents/skills
cp -R /tmp/benchflow-ai__skillsbench/registry/terminal_bench_2.0/letta_skills_batch/terminal_bench_2_0_gpt2-codegolf/environment/skills/gpt2-codegolf ~/.claude/skills/gpt2-codegolf # Claude Code
cp -R /tmp/benchflow-ai__skillsbench/registry/terminal_bench_2.0/letta_skills_batch/terminal_bench_2_0_gpt2-codegolf/environment/skills/gpt2-codegolf ~/.agents/skills/gpt2-codegolf # Codex, Cursor, Gemini CLI, Copilot, OpenCode