lowmason/agent-skills
Portable AI agent skills and customization configs for Claude, Codex, and Gemini
field: Developer tools workflow: Planning & specsPR / code reviewGit & commitsRefactoringTesting & TDD
stack: Python
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 lowmason/agent-skills/.claude ./rig-agent-skills # 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 lowmason/agent-skills/.claude/skills/cc-guide .claude/skills/cc-guide $ npx degit lowmason/agent-skills/archive/skills/bls-data-context .claude/skills/bls-data-context $ npx degit lowmason/agent-skills/archive/skills/classification-codes .claude/skills/classification-codes $ npx degit lowmason/agent-skills/archive/skills/geographic-codes .claude/skills/geographic-codes $ npx degit lowmason/agent-skills/skills/bayesian-workflow .claude/skills/bayesian-workflow $ npx degit lowmason/agent-skills/skills/brainstorming .claude/skills/brainstorming $ npx degit lowmason/agent-skills/skills/clean-code .claude/skills/clean-code $ npx degit lowmason/agent-skills/skills/clean-coder .claude/skills/clean-coder $ npx degit lowmason/agent-skills/skills/creative-thinking .claude/skills/creative-thinking $ npx degit lowmason/agent-skills/skills/deep-learning .claude/skills/deep-learning $ npx degit lowmason/agent-skills/skills/derive-roadmap .claude/skills/derive-roadmap $ npx degit lowmason/agent-skills/skills/describe-critique-methodology .claude/skills/describe-critique-methodology $ npx degit lowmason/agent-skills/skills/design-architecture .claude/skills/design-architecture $ npx degit lowmason/agent-skills/skills/develop-testing-strategy .claude/skills/develop-testing-strategy $ npx degit lowmason/agent-skills/skills/dispatching-parallel-agents .claude/skills/dispatching-parallel-agents $ npx degit lowmason/agent-skills/skills/document-python-package .claude/skills/document-python-package $ npx degit lowmason/agent-skills/skills/evaluate-deep-learning .claude/skills/evaluate-deep-learning $ npx degit lowmason/agent-skills/skills/executing-plans .claude/skills/executing-plans $ npx degit lowmason/agent-skills/skills/explore-data .claude/skills/explore-data $ npx degit lowmason/agent-skills/skills/finishing-a-development-branch .claude/skills/finishing-a-development-branch $ npx degit lowmason/agent-skills/skills/llm-wiki .claude/skills/llm-wiki $ npx degit lowmason/agent-skills/skills/optimize-jax .claude/skills/optimize-jax $ npx degit lowmason/agent-skills/skills/prepare-handoff .claude/skills/prepare-handoff $ npx degit lowmason/agent-skills/skills/receiving-code-review .claude/skills/receiving-code-review $ npx degit lowmason/agent-skills/skills/recommend-causal-design .claude/skills/recommend-causal-design $ npx degit lowmason/agent-skills/skills/recommend-probabilistic-model .claude/skills/recommend-probabilistic-model $ npx degit lowmason/agent-skills/skills/recommend-visualization .claude/skills/recommend-visualization $ npx degit lowmason/agent-skills/skills/requesting-code-review .claude/skills/requesting-code-review $ npx degit lowmason/agent-skills/skills/subagent-driven-development .claude/skills/subagent-driven-development $ npx degit lowmason/agent-skills/skills/systematic-debugging .claude/skills/systematic-debugging $ npx degit lowmason/agent-skills/skills/tech-debt .claude/skills/tech-debt $ npx degit lowmason/agent-skills/skills/test-driven-development .claude/skills/test-driven-development $ npx degit lowmason/agent-skills/skills/track-model-experiments .claude/skills/track-model-experiments $ npx degit lowmason/agent-skills/skills/tune-hyperparameters .claude/skills/tune-hyperparameters $ npx degit lowmason/agent-skills/skills/using-git-worktrees .claude/skills/using-git-worktrees $ npx degit lowmason/agent-skills/skills/validate-data .claude/skills/validate-data $ npx degit lowmason/agent-skills/skills/verification-before-completion .claude/skills/verification-before-completion $ npx degit lowmason/agent-skills/skills/write-with-asd-ste100 .claude/skills/write-with-asd-ste100 $ npx degit lowmason/agent-skills/skills/writing-plans .claude/skills/writing-plans $ npx degit lowmason/agent-skills/skills/writing-skills .claude/skills/writing-skills
$ curl -fsSL --create-dirs -o .claude/agents/cc-guide-verifier.md https://raw.githubusercontent.com/lowmason/agent-skills/main/.claude/agents/cc-guide-verifier.md $ curl -fsSL --create-dirs -o .claude/agents/code-reviewer.md https://raw.githubusercontent.com/lowmason/agent-skills/main/agents/code-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/debugger.md https://raw.githubusercontent.com/lowmason/agent-skills/main/agents/debugger.md $ curl -fsSL --create-dirs -o .claude/agents/docs-writer.md https://raw.githubusercontent.com/lowmason/agent-skills/main/agents/docs-writer.md $ curl -fsSL --create-dirs -o .claude/agents/explore.md https://raw.githubusercontent.com/lowmason/agent-skills/main/agents/explore.md $ curl -fsSL --create-dirs -o .claude/agents/security-auditor.md https://raw.githubusercontent.com/lowmason/agent-skills/main/agents/security-auditor.md $ curl -fsSL --create-dirs -o .claude/agents/task-reviewer.md https://raw.githubusercontent.com/lowmason/agent-skills/main/agents/task-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/test-runner.md https://raw.githubusercontent.com/lowmason/agent-skills/main/agents/test-runner.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 16.2k rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(./.env.*)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(git push --force:*)",
"Read(**/*.pem)",
"Read(.env)",
"Read(~/.aws/**)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Bash(rm -rf *)",
"Bash(sudo *)",
"Bash(sudo:*)",
"Bash(rm -rf /)",
"Bash(git push -f:*)",
"Read(.env.*)",
"Bash(git reset --hard*)",
"Bash(rm -rf /*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(gh pr merge *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(gh pr merge:*)",
"Bash(gh pr create *)",
"Bash(git reset --hard:*)",
"Bash(npm publish:*)",
"Bash(git rebase *)"
]
}
} Skills (40)
cc-guidebls-data-contextclassification-codesgeographic-codesbayesian-workflowbrainstormingclean-codeclean-codercreative-thinkingdeep-learningderive-roadmapdescribe-critique-methodologydesign-architecturedevelop-testing-strategydispatching-parallel-agentsdocument-python-packageevaluate-deep-learningexecuting-plansexplore-datafinishing-a-development-branchllm-wikioptimize-jaxprepare-handoffreceiving-code-reviewrecommend-causal-designrecommend-probabilistic-modelrecommend-visualizationrequesting-code-reviewsubagent-driven-developmentsystematic-debuggingtech-debttest-driven-developmenttrack-model-experimentstune-hyperparametersusing-git-worktreesvalidate-dataverification-before-completionwrite-with-asd-ste100writing-planswriting-skills
Subagents (8)
| cc-guide-verifier model: sonnet | Checks this repo's Claude Code guide against the docs, from a packet the /cc-guide skill builds. Only /cc-guide dispatches it. |
| code-reviewer model: opus | Use for reviews dispatched by the requesting-code-review skill and for subagent-driven-development's final whole-branch review; per-task reviews go to the task-reviewer agent. Read-only code reviewer |
| debugger model: sonnet | Fixes one self-contained, reproducible failure in an isolated context — a named failing test or a crashing script where the dispatch carries everything needed to reproduce. Reproduces first, isolates |
| docs-writer model: sonnet | Writes technical documentation in an isolated context — repo/package READMEs, analysis writeups for finished data work, docstrings and API docs, and general technical docs (runbooks, guides, ADR prose |
| Explore model: haiku | Read-only search agent for broad fan-out searches — locates code across many files, directories, and naming conventions, and reports path:line references with one-line relevance notes rather than file |
| security-auditor model: opus | Use for a security-focused review of a diff, branch, or repo — injection risks, committed secrets and credential handling, insecure deserialization, TLS verification, dependency risks. Read-only; repo |
| task-reviewer model: sonnet # task-review floor per subagent | Read-only task-scoped reviewer that checks one task's implementation against its brief — spec compliance first, then code quality. Dispatched by the subagent-driven-development skill after each task; |
| test-runner model: haiku | Runs one test suite in isolation and reports results without polluting the caller's context. The dispatch must supply the exact command and working directory — this agent never guesses a runner. Repor |
Hooks (2)
| event | matcher | runs |
|---|---|---|
| SessionStart | startup | "$CLAUDE_PROJECT_DIR"/build/cc_guide/notice.sh show |
| SessionStart | startup | "$CLAUDE_PROJECT_DIR"/build/cc_guide/notice.sh refresh |
Slash commands (3)
/deferred/fix-issue/license-audit
Similar rigs
minhtuluc/Claude-Code-Configs
—
Orchestrator 2.2k tok · · Q15
sodiumjoe/dotfiles
dotfiles
Skill Collector 1.1k tok · · Q3
RichardBray/dotfiles
My personal configuration files and settings
Skill Collector 671 tok · · Q0
izyanrajwani/agent-skills-library
🛠️ Enhance programming workflows with a curated library of agent skills for tasks like code review, debugging, and planning for efficient project execution.
Skill Collector 854 tok · · Q0