agentleFS
917Dhj917Dhj/dkitchen/skills/setup-dkitchen/SKILL.md

setup-dkitchen

Skill10 starsChanged 8 days agoMITView on GitHub

Install the dkitchen multi-agent workflow into this repo. Finds out what the repo and the user need, then walks through each build skill in turn. Run once per repo, or again to change the setup.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read the repo's starting state.

    line 15
  2. Distinctive:

    Needs: Ask each as a yes or no:

    line 34
  3. Distinctive:

    Run the checks in Installing, then each chosen build skill's own verify step.

    line 53

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Set up dkitchen

Install the kitchen into the current repo. The user is the head chef: every rule you install exists to keep their attention on direction and decisions. This skill is the entry: it checks the prerequisites, settles the language and what the user needs, then walks through each build skill the user needs. Each build skill settles, drafts, and writes its own part. You verify and hand over at the end.

Shared install steps, such as prerequisites, language, rendering, and pointer blocks, are in kitchen/README.md. Read it first.

1. Explore

Read the repo's starting state; assume nothing.

  • Prerequisites: each one in Installing.
  • Existing setup: AGENTS.md, CLAUDE.md, .claude/CLAUDE.md, docs/agents/, .dkitchen.toml, .claude/settings.json. Note which build skills already ran: docs/agents/roles.md for build-team, docs/agents/code.md for build-repo, docs/agents/runner.md for build-runner, docs/agents/unattended.md for build-night.
  • The work: the README, the main code, how tests run, whether scripts loop over items, and whether there are long jobs, several parallel workstreams, or a user who wants progress while away.
  • Language: the language the user writes to you in, and the language of the repo's docs.

Done when you can name, with evidence, each prerequisite's state, which build skills already ran, and a recommendation for each need in step 3.

2. Present the findings

Summarize what is present and what is missing in a short list. Missing prerequisites come first, each with the command that fixes it. When Claude Code or Codex is unavailable, say plainly that this version of dkitchen needs both, and ask whether to stop or to install the rules anyway.

3. Settle the language and the needs

One question at a time. Lead each with your recommendation so the user can accept it in a word, and skip any that exploration already settled.

Language. Recommend the language the user writes to you in, and settle the vocabulary as Installing says. The build skills reuse this answer and do not ask again.

Needs. Ask each as a yes or no:

NeedRecommendBuild skill
A team of agents: roles, tickets, PR reviewAlways yes; every other part builds on itbuild-team
Code the agents write and changeYes for any repo with code the agents change; the reference checks need Pythonbuild-repo
Work done item by item: evaluations, batch inference, data processingYes when the repo has scripts that loop over items; needs build-repobuild-runner
Work that keeps moving while the user is awayYes when the repo has long jobs or several parallel workstreams, or the user asks for itbuild-night

Done when the language is settled and every need has an answer.

4. Build each part

Take the chosen build skills in the table's order: open each one's SKILL.md and follow it from its first step to its write step, carrying the settled language into it. A build skill already installed is updated in place. Finish one before starting the next, since each later skill builds on the earlier ones' files.

Done when every chosen build skill has written its approved draft.

5. Verify

Run the checks in Installing, then each chosen build skill's own verify step. Fix what fails.

Done when every check passes, or the user has been told exactly which one fails and why.

6. Hand over

Give the user each build skill's hand-over points in one short list, in the table's order. End with: the rules are theirs to edit, and /improve-kitchen, run every couple of weeks, turns what keeps going wrong into checks, tools, or rules; each build skill can be run again on its own to change its part; and /setup-dkitchen runs the whole setup again.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.