Claude Code Research Engineer Skill for Algorithm Comparisons
Define time, memory, and correctness constraints before comparing candidate algorithms. Turn performance assumptions into checks that can be reproduced.
中文Install in Claude Code
Use Claude Code with the pinned skill directory. Bring a well-defined problem, input sizes, and sample cases for comparing correctness.
- Get the complete `cli-tool/components/skills/ai-research/research-engineer` directory from the pinned GitHub checkout.
- Copy it under the target project's `.claude/skills/`. Confirm `.claude/skills/research-engineer/SKILL.md`, open Claude Code at the project root, and call `/research-engineer`.
Example task
Compare pairwise merging and heap-based merging of sorted integer lists. State total element count N and list count k. Include empty lists, duplicates, and unequal lengths in correctness checks.What to check
Expect applicability and complexity notes for both approaches, along with cases that confirm ordering and preservation of elements. Leave benchmark results for an actual run.
Original SKILL.md on GitHub · Updated
Frequently asked question
How do I keep complexity analysis separate from benchmark results?
Label theoretical bounds, the benchmark environment, the dataset, and measured metrics separately. Mark measurements that have yet to run as pending.