center-for-humans-and-machines/claude-code-template
Claude Code research project template — agents and workflows for engineering, analysis, manuscript writing, and presentations
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
Copy this rig
# review before running: this installs third-party code
$ npx degit center-for-humans-and-machines/claude-code-template/.claude ./rig-claude-code-template # inspect, then merge into .claude/ MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ npx degit center-for-humans-and-machines/claude-code-template/.claude/skills/analyse .claude/skills/analyse $ npx degit center-for-humans-and-machines/claude-code-template/.claude/skills/cluster .claude/skills/cluster $ npx degit center-for-humans-and-machines/claude-code-template/.claude/skills/commit .claude/skills/commit $ npx degit center-for-humans-and-machines/claude-code-template/.claude/skills/engineer .claude/skills/engineer $ npx degit center-for-humans-and-machines/claude-code-template/.claude/skills/ms-pdf-extract .claude/skills/ms-pdf-extract $ npx degit center-for-humans-and-machines/claude-code-template/.claude/skills/pr .claude/skills/pr $ npx degit center-for-humans-and-machines/claude-code-template/.claude/skills/share-ready .claude/skills/share-ready
$ curl -fsSL --create-dirs -o .claude/agents/analysis-analyst.md https://raw.githubusercontent.com/center-for-humans-and-machines/claude-code-template/main/.claude/agents/analysis-analyst.md $ curl -fsSL --create-dirs -o .claude/agents/analysis-reviewer.md https://raw.githubusercontent.com/center-for-humans-and-machines/claude-code-template/main/.claude/agents/analysis-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/engineer-fast.md https://raw.githubusercontent.com/center-for-humans-and-machines/claude-code-template/main/.claude/agents/engineer-fast.md $ curl -fsSL --create-dirs -o .claude/agents/engineer-max.md https://raw.githubusercontent.com/center-for-humans-and-machines/claude-code-template/main/.claude/agents/engineer-max.md $ curl -fsSL --create-dirs -o .claude/agents/engineer.md https://raw.githubusercontent.com/center-for-humans-and-machines/claude-code-template/main/.claude/agents/engineer.md $ curl -fsSL --create-dirs -o .claude/agents/experiment-designer.md https://raw.githubusercontent.com/center-for-humans-and-machines/claude-code-template/main/.claude/agents/experiment-designer.md $ curl -fsSL --create-dirs -o .claude/agents/ms-brainstorm.md https://raw.githubusercontent.com/center-for-humans-and-machines/claude-code-template/main/.claude/agents/ms-brainstorm.md $ curl -fsSL --create-dirs -o .claude/agents/ms-figure-creator.md https://raw.githubusercontent.com/center-for-humans-and-machines/claude-code-template/main/.claude/agents/ms-figure-creator.md $ curl -fsSL --create-dirs -o .claude/agents/ms-literature-researcher.md https://raw.githubusercontent.com/center-for-humans-and-machines/claude-code-template/main/.claude/agents/ms-literature-researcher.md $ curl -fsSL --create-dirs -o .claude/agents/ms-writer.md https://raw.githubusercontent.com/center-for-humans-and-machines/claude-code-template/main/.claude/agents/ms-writer.md $ curl -fsSL --create-dirs -o .claude/agents/planner.md https://raw.githubusercontent.com/center-for-humans-and-machines/claude-code-template/main/.claude/agents/planner.md $ curl -fsSL --create-dirs -o .claude/agents/presentation-creator.md https://raw.githubusercontent.com/center-for-humans-and-machines/claude-code-template/main/.claude/agents/presentation-creator.md $ curl -fsSL --create-dirs -o .claude/agents/reviewer.md https://raw.githubusercontent.com/center-for-humans-and-machines/claude-code-template/main/.claude/agents/reviewer.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 7,204 rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(rm -rf *)",
"Read(**/*.pem)",
"Bash(rm -rf /)",
"Bash(git push --force:*)",
"Bash(sudo *)",
"Read(.env)",
"Bash(rm -rf /*)",
"Read(./.env.*)",
"Read(~/.aws/**)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(sudo:*)",
"Bash(git reset --hard*)",
"Bash(git reset --hard:*)",
"Read(.env.*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(rm:*)",
"Bash(git rebase *)",
"Bash(wget *)",
"Bash(npm publish:*)",
"Bash(git commit *)",
"Bash(gh pr merge *)"
]
}
} Skills (7)
Subagents (13)
| analysis-analyst model: opus | Use this agent to implement data analysis, processing pipelines, visualisations, or statistical models. Handles reading, editing, and writing code to fulfil a task defined by the user or a plan. |
| analysis-reviewer model: opus | Use this agent to critically review analysis reports. Evaluates informativeness of each section and visualisation, flags uninformative content for removal, and suggests targeted improvements. Does not |
| engineer-fast model: sonnet | Quick implementation tasks. Lower effort, faster turnaround. Good for mechanical changes, simple scripts, file moves. |
| engineer-max model: opus | Maximum effort implementation. For critical, complex, or tricky code where correctness matters most. |
| engineer model: opus | Implement features, fix bugs, write code. High effort — reads context carefully, tests, commits. |
| experiment-designer model: opus | Use this agent to design experiments by creating YAML config files and writing hypotheses. It reads existing configs and code to understand the config schema, then produces correct, self-contained con |
| ms-brainstorm model: opus | >- |
| ms-figure-creator model: opus | Generate and refine conceptual figures for the manuscript using OpenAI's gpt-image-2. The agent iterates: generate → visually inspect → critique → refine. Use when the user wants a new conceptual diag |
| ms-literature-researcher model: opus | Use this agent to perform literature research — find relevant papers, summarize findings, identify related work, compile references, and investigate specific claims. Scales effort to the question — fu |
| ms-writer model: opus | Write or rewrite manuscript prose (LaTeX). Use when the user wants to turn outline notes or bullet drafts into polished text, or rewrite a section. |
| planner model: opus | Plan and document next steps. Creates plans in doc/plans/, defines task breakdowns. Never writes implementation code. |
| presentation-creator model: opus | Use this agent when the user wants to create a presentation, slide deck, or structured visual content. This includes requests to build presentations from scratch, convert notes or documents into prese |
| reviewer model: opus | Review code for correctness, bugs, and unnecessary complexity. Does not write code — only reports findings. |
Similar rigs
ysoftman/myenv
personal environment settings
Skill Collector 2.0k tok ·
Windy3f3f3f3f/claude-code-from-scratch
Build your own Claude Code from scratch. 🔍 Claude Code 开源了 50 万行代码，读不动？用 ~5000 行 TypeScript / Python 从零复现核心架构，11 章分步教程带你理解 coding agent 精髓
Pragmatist 2.6k tok ·
ramarlina/agx
agx: Run AI coding agents as a persistent team with objectives, memory, and coordinated work. The same agents built this tool — 167+ merged PRs, 93% clean.
Minimalist 167 tok ·
jasonlong/dotfiles
Files. With dots in front of them.
Pragmatist 336 tok ·