claudefiles
take everything here as a proof of concept, i just need a place to centralize so i can pick the best lol
Iβll try to make it obvious what is or isnt human written, but definitely at least assume its all slop until proven/told otherwise, thanks
Feel free to ask any questions or suggestions tho, and, if you see any slop you want to fix, by all means, Iβll probably accept it, but this isnt my master user level .claude or anything, at least not yet, thats actually mostly empty, hereβs why:
most of this this stuff is just a few prompts to make, and i was lazy and figured it was better to just keep remaking them project level, so i can collect them and compare. thatβs what this is for, or will be for, at the moment
i have no idea which oneβs the best, and im gonna have to β¦
claudefiles
take everything here as a proof of concept, i just need a place to centralize so i can pick the best lol
Iβll try to make it obvious what is or isnt human written, but definitely at least assume its all slop until proven/told otherwise, thanks
Feel free to ask any questions or suggestions tho, and, if you see any slop you want to fix, by all means, Iβll probably accept it, but this isnt my master user level .claude or anything, at least not yet, thats actually mostly empty, hereβs why:
most of this this stuff is just a few prompts to make, and i was lazy and figured it was better to just keep remaking them project level, so i can collect them and compare. thatβs what this is for, or will be for, at the moment
i have no idea which oneβs the best, and im gonna have to resolve naming collisions probably, but, i think itβs safe to say iβm already over-engineering the canonical truth of slop, so, enjoy!
Whatβs Here
Claude Code configuration files: agents, commands, hooks, and skills for software development workflows.
Structure
claudefiles/
βββ agents/
β βββ general/ (6 agents: git-ops, github-cli, web-search, etc.)
β βββ golang/ (5 agents: architect, explorer, implementer, optimizer, reviewer)
β βββ python/ (4 agents: code-writer, code-pedant, test-runner, slash-command-architect)
β βββ typescript/ (5 agents: craftsman, code-reviewer, test-architect, test-runner, quick-impl)
β βββ docs/ (1 agent: docs-architect)
βββ commands/
β βββ docs/ (generate-overview.md)
β βββ golang/ (implement.md)
β βββ typescript/ (implement.md)
βββ hooks/
β βββ golang/ (go-fmt.sh, go-vet.sh, go-precommit.sh)
β βββ security/ (check-secrets.py)
βββ skills/
βββ golang/ (6 categories with nested sub-skills)
βββ concurrency/ (channels, context, goroutines, sync)
βββ errors/ (checking, sentinel, wrapping)
βββ interfaces/ (design, embedding, pollution)
βββ linting/ (golangci, staticcheck, vet)
βββ nil/ (interface, map, pointer, slice)
βββ testing/ (benchmarks, helpers, subtests, table)
Components
agents/: Specialized AI assistants by language/domain. general/ contains cross-language utilities (git, GitHub, web search). Language dirs contain implementation, review, and testing agents.
commands/: Slash commands for workflow automation (/golang/implement, /typescript/implement, /docs/generate-overview)
hooks/: Event-driven scripts that auto-run on tool use. golang/ runs go fmt/go vet on file changes. security/ scans for leaked secrets.
skills/: Reference knowledge as nested SKILL.md files. Currently Go patterns only.
Usage
Copy to .claude/ at project or user level. See CLAUDE.md for the recommended parallel wave workflow.