Coding Agent Workflows

coding-agent-workflows

Coding standards, agent roles, skills, and multi-step workflows that read the same whether you drive Claude Code, Codex, Amp, or anything that reads an `AGENTS.md`. One neutral source renders to a native config for each agent, and the agent-specific pieces stay scoped to the agents that need them.

Everything is pre-rendered and committed. There is no build step between clone and use, and installing is a plain file copy: no hooks, no daemons, nothing running in the background.

Install

git clone https://github.com/sjarmak/coding-agent-workflows.git
cd coding-agent-workflows

Then install for your agent:

./install.sh claude          # Claude Code config into ./.claude (project-level; `./install.sh claude ~` for user-level)
./install.sh codex           # AGENTS.md + AGENTS.full.md into the current dir, Codex config into ~/.codex
./install.sh agents          # AGENTS.md (thin index) + AGENTS.full.md (Amp, Aider, Gemini CLI, ...)
./install.sh init            # scaffold a thin, project-specific AGENTS.md (intention + pointers)
./install.sh upgrade         # re-install, then prune files dropped since the last install
./install.sh remove          # delete exactly what a prior claude install placed in .claude

Pass a destination as the second argument to target a specific project, for example `./install.sh claude ~/work/myrepo`. The `claude` install writes a manifest of every file it owns and backs up any pre-existing file it would otherwise overwrite; `remove` undoes exactly that manifest, so a user-level install never silently clobbers hand-authored config. A separate `fleet` subcommand installs a machine-level conformance scanner into `~/.claude/fleet`; it is the one mode that wires a hook, and it says so when it runs.

If you would rather not run a script, copy what you need: Claude Code reads a `.claude/` directory, so copy `targets/claude/{rules,agents,skills,commands}` into one. Codex reads `AGENTS.md` at the rep