agentleFS
M-BajalanM-Bajalan/kibsu-lab/skills/architect/SKILL.md

architect

Skill0 starsChanged 32 days agoMITView on GitHub

The conventions this repository enforces: what every file must declare, where it belongs, and the commands that prove a new or moved file follows the rules before it lands. Run it whenever you add, move, or rename anything.

174 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use ecosystem-reference, fractal-delegation and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • naman10parikhSkill

    ecosystem-reference

    Quick-access directory of Claude Code ecosystem repos, skills, plugins, MCPs, and patterns. Use when improving the harness.

    Also in 5 projects that use this

  • naman10parikhSkill

    fractal-delegation

    Also in 5 projects that use this

  • tristanphvnSkill

    ketchup-video

    Catch the user up like /keel:ketchup, with a short silent video of slides for what happened and the waiting calls as text under it. Use for /keel:ketchup-video.

    Also in 14 projects that use this

  • jaydubya818Skill

    figure-it-out

    Use when novel or multi-phase work needs an auditable workflow and no narrower playbook fits.

    Also in 55 projects that use this

  • ianclearySkill

    principle-explain-the-number

    Apply before you trust, report, or act on a number you measured: a speedup, a regression, a throughput, a latency, or an eval result. Find what limits it, and rule out that it measured something other than the work you think.

    Also in 29 projects that use this

  • eryzerzSkill

    make-bot-ui

    Use when building a custom UI (page, dashboard, buttons) that should wake a Grok Bot over a webhook, when the user must provide a webhook sender key, or when exposing that UI on Tailscale.

    Also in 20 projects that use this

  • +6 more from naman10parikh

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 2 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 162 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Every document under docs/ carries frontmatter with a type: field.

    line 19
  2. Very specific:

    Read the line genre declared in frontmatter: N unit(s), 0 disagreeing with detection.

    line 36
  3. Distinctive:

    Never open a second home for a fact that already has one.

    line 55
  4. Distinctive:

    If you cannot decide whether a change is worth a CHANGELOG entry, write the entry.

    line 90

“Common” is measured: kept by at least 6 in 10 of the 162 edited versions of this file we could read. “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

architect

Conventions only survive if a command can check them. Everything below names the command that proves it, because a rule nobody can verify is a preference, and a preference decays into an argument the next time two agents disagree.

Rule 1 -- every file declares what it is

Every document under docs/ carries frontmatter with a type: field. Every skill carries name, type: skill, and genre:. Nothing lands without its declaration.

python -m kibsu report .

The Know your conventions line reads consistent frontmatter (type) - enforceable while this holds. If it says anything else, a file you added is missing its type: -- find it before you commit, not after.

Rule 2 -- declare the genre honestly

python -m kibsu audit .

Read the line genre declared in frontmatter: N unit(s), 0 disagreeing with detection. The 0 is the number that matters. A non-zero count means a skill claims one genre while its text reads as another.

A doctrine unit scoring near 0% checkable is correct and expected -- judgment guidance has nothing to check. Relabelling a doctrine unit as procedure would raise the headline number without changing a single instruction, which is the exact deception this repo's tooling exists to catch. Declare what the file is; let the number land where it lands.

Rule 3 -- one fact, one home

The factIts only home
What is true right nowdocs/STATE.md
What already happeneddocs/CHANGELOG.md
A reusable lessondocs/memory/learnings/<slug>.md
What a multi-step job will dodocs/plans/<date>-<slug>.md
Where roots and gates are wired.kibsu.json

Never open a second home for a fact that already has one. A duplicated fact does not stay duplicated -- one copy gets updated, the other rots, and the next agent cannot tell which one is lying. When you are tempted to restate something, link to it instead.

Rule 4 -- where things belong

PathHolds
docs/documents, each with type: frontmatter
docs/plans/plan files (skills/planner/SKILL.md)
docs/memory/learnings/one lesson per file
skills/<name>/SKILL.mdone skill per directory
tools/runnable Python, run from the repo root (check_task.py takes a task id)
tasks/task briefs graded by tools/check_task.py
data/generated, never committed -- see PROVENANCE.md

A file placed outside this table needs the table updated in the same commit, or it needs to move. Do not leave the exception undocumented.

Rule 5 -- rebuild the index after you add or move a file

The index is what lets an agent find your docs at all. It is gitignored, so it is never carried in by a clone and never updated by anyone else's commit.

python -m kibsu index .
python -m kibsu report .

report reads Find your docs from that index. A file that exists but is not indexed is invisible to every check that follows, so an index rebuild you skipped reads exactly like a file you never wrote.

Rule 6 -- when unsure, record more, not less

If you cannot decide whether a change is worth a CHANGELOG entry, write the entry. A too-small entry is edited in seconds; a missing one is history nobody can reconstruct. The same holds for a learning you are unsure is reusable -- the cost of writing it is a minute, and the cost of re-deriving it is however long it took the first time.

What counts as a change worth recording

A change is worth recording if it would confuse the next agent had it gone unrecorded. Editing a tool, a skill, a task, or a config qualifies. Reading, auditing, or answering a question does not -- but if the answer took real work to find, that answer is a learning under Rule 3, not nothing.

Checklist before you say "it follows the conventions"

  • Does every file you added or edited carry its type: frontmatter (and name + genre: if it is a skill)?
  • Did python -m kibsu report . still print consistent frontmatter (type) - enforceable?
  • Did python -m kibsu audit . print 0 disagreeing with detection?
  • Is every fact you wrote in exactly one home, with links rather than copies?
  • Is every file you added covered by the path table in Rule 4, or was the table updated?
  • Did you re-run python -m kibsu index . after adding or moving a file?
  • Did you log the change per skills/change-log/SKILL.md?

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.