research-engineer — Claude Code skill: what it does, how to install
Turns Claude Code into a strict research engineer focused on scientific rigor and complete runnable code without placeholders.
npx claude-code-templates@latest --skill ai-research/research-engineerWhat research-engineer does
This skill enforces uncompromising academic standards in Claude Code. It eliminates conversational fluff and starts directly with technical criticism. If an approach breaks theoretical constraints, such as attempting to parse nested structures using regular expressions, the agent rejects the premise before writing any logic.
Code output avoids shortcuts. The skill prohibits comments like // insert logic here. If a thread-safe solution requires hundreds of lines of boilerplate, the agent writes every line. Algorithms include formal assertions or property-based tests using frameworks like Hypothesis. When hitting token caps, the model halts with an explicit marker to continue cleanly in the next turn.
Tool selection depends strictly on computational requirements. Rather than defaulting to general-purpose languages, it directs numerical work toward Julia or Fortran and memory-critical infrastructure toward Rust or Ada.
What it helps with
- evaluate algorithmic time and space complexity against theoretical limits
- generate full compilable implementations without omitted logic
- select appropriate programming languages for high-performance computing or formal proofs
- identify theoretical flaws in system design before implementation
Example prompts for the agent
- Analyze the asymptotic complexity of this recurrence relation and prove the lower bound.
- Implement a lock-free queue in C++20 using explicit memory ordering semantics.
- Critique this distributed consensus architecture and outline failure modes.
How to install research-engineer
1claude-code-templates
npx claude-code-templates@latest --skill ai-research/research-engineer2Skills CLI (Vercel)
npx skills add davila7/claude-code-templates --skill research-engineer -a claude-code3Manually: put the files in place
~/.claude/skills/research-engineer/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/research-engineer/
Manually: put the files in place
.agents/skills/research-engineer/SKILL.mdFor all projects use ~/.agents/skills/
Manually: put the files in place
.cursor/skills/research-engineer/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install davila7/claude-code-templates research-engineer2Manually: put the files in place
.github/skills/research-engineer/SKILL.mdPlatform CLI
gemini skills install https://github.com/davila7/claude-code-templates --path cli-tool/components/skills/ai-research/research-engineerManually: put the files in place
.devin/skills/research-engineer/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about research-engineer
How to install research-engineer in Claude Code?
Run npx claude-code-templates@latest --skill ai-research/research-engineer in your terminal.
What does research-engineer do?
It configures Claude Code to act as an uncompromising academic researcher who rigorously evaluates premises, enforces computational correctness, and delivers complete code.
Does this skill allow code placeholders?
No. The operational mandate explicitly forbids incomplete code and requires fully functional implementations without skipped blocks.