agentleFS
protonium-labsprotonium-labs/axiomcore-plugin/skills/setup/SKILL.md

setup

Skill3 starsChanged 3 months agoMITView on GitHub

Owns the setup/ area and makes the system adaptable. Actions - new (draft a capability setup file via guided interview), run (execute a pending setup file), list, overview. Also owns structural migrations - changing the hierarchy depth of a project or workstation. First-run onboarding is owned by /axiom, not here.

191 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use add-voice-transcription, add-discord 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.
  • juancruzmunozalbeloSkill

    add-voice-transcription

    Add voice message transcription to NanoClaw using OpenAI's Whisper API. Automatically transcribes WhatsApp voice notes so the agent can read and respond to them.

    Also in 5 projects that use this

  • MicrosoftSkill

    add-discord

    Add Discord bot channel integration via Chat SDK.

    Also in 11 projects that use this

  • MicrosoftSkill

    add-slack

    Add Slack channel integration via Chat SDK.

    Also in 11 projects that use this

  • juancruzmunozalbeloSkill

    add-telegram-swarm

    Add Agent Swarm (Teams) support to Telegram. Each subagent gets its own bot identity in the group. Requires Telegram channel to be set up first (use /add-telegram). Triggers on "agent swarm", "agent teams telegram", "telegram swarm", "bot pool".

    Also in 7 projects that use this

  • +8 more from microsoft

The file

We split this skill into the rules it gives an agent

8 rules. 0 are common, 5 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 171 edited versions we read keep these.

  • 5

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If the requested action is ambiguous or contradictory, do NOT guess: flag it and ask for clarification before executing.

    line 12
  2. Distinctive:

    If the action was unknown, say so first, then show the table.

    line 31
  3. Distinctive:

    Check $ARGUMENTS (after "new"): the capability idea.

    line 39
  4. Distinctive:

    Read the draft back to the user.

    line 48
  5. Very specific:

    Check $ARGUMENTS (after "run"): match against files in setup/ (not done/).

    line 54
  6. Distinctive:

    Read the file. Verify prerequisites.

    line 55
  7. Very specific:

    Update the Last updated date in every file changed.

    line 92
  8. Very specific:

    Never overwrite a filled soul.md or brand-config.md without explicit confirmation — those are owned by /axiom.

    line 96

“Common” is measured: kept by at least 6 in 10 of the 171 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

Task

Execute the requested action: $ARGUMENTS

Calling rule: /setup <action> <arguments>.

If the requested action is ambiguous or contradictory, do NOT guess: flag it and ask for clarification before executing.

First-run onboarding and module activation are owned by /axiom — if the user asks for initial setup, point them to /axiom start.


If action is empty

Describe this skill: its purpose and the actions table from "overview". Do not execute anything.

If action is "overview" (or unknown)

ArgumentAction
new <capability idea>Guided interview + structure analysis → draft a new setup file in setup/
run <name>Execute a pending setup file and archive it to setup/done/
listShow pending and installed capabilities
overviewThis table

If the action was unknown, say so first, then show the table.


If action is "new"

Goal: turn a capability idea into a reviewable setup file — nothing is implemented in this action.

  1. Check $ARGUMENTS (after "new"): the capability idea. If empty: ask what capability the user wants. If the idea is a hierarchy-depth change for an existing project or workstation, skip the capability interview and follow "Structural migrations" below instead — the file has a fixed content list.
  2. Interview, one block at a time, only what the idea leaves open:
    • Goal: what should the capability do, for whom, how often (recurring or one-off)?
    • Inputs: what arrives, where does it land (folder, MCP, email)?
    • Outputs: what is produced, where does it go?
    • Integrations: which MCPs/services are needed?
    • Additions: new skills, agents, vault areas?
  3. Analyze the current structure: root CLAUDE.md rules (hierarchy, WP-only work, vault mirroring), existing projects, vault/index.md, installed capabilities in setup/done/. The capability must fit the system rules — e.g. a "drop folder for bills" becomes a WP resources/ folder inside a finance project, not a foreign folder.
  4. Draft setup/su-<NNN>-<name>.md (kebab-case name, no -setup suffix — the folder and the acronym already say it): sections Capability, Prerequisites, Installs (concrete list), Open decisions (what remains to clarify at run time), Definition (detailed behavior). Status: pending. Allocate <NNN> here, at drafting, from Next setup in setup/README.md and bump the counter — monotonic, never reused; a discarded draft does not free its number. Setup files are deliverables: the file carries the standard header block and changelog (.claude/templates/output-document-template.md, acronym SU) and keeps its name when it is later archived to done/ — the number is how it is cited afterwards.
  5. Read the draft back to the user; iterate until approved. Lint the file.

If action is "run"

  1. Check $ARGUMENTS (after "run"): match against files in setup/ (not done/). No match or empty: run "list" and ask which one.
  2. Read the file. Verify prerequisites; if one is missing (e.g. MCP not configured), stop and tell the user what to do first.
  3. Clarify all open decisions with the user.
  4. Implement what "Installs" defines. Workflow discipline applies: present the implementation plan from the file, get approval, execute step by step. For a hierarchy-change file the folder map and registry rewrite take the place of "Installs" — follow the execution rules in "Structural migrations" below.
  5. Record the decisions and results in the setup file, bump its version and add a changelog row, set Status: installed with date, then move it to setup/done/ under its own name — archiving never renames a setup file.
  6. Lint all changed .md files.

If action is "list"

Show two tables: Pending (files in setup/, excluding README) and Installed (files in setup/done/, with installation date from the file). One line per capability: name, what it installs, prerequisites state if known.


Structural migrations (hierarchy depth)

/project and /ws refuse to change the hierarchy depth of an existing unit and send the user here. Such a change renumbers folders and — on the project side — rewrites every task ID in the project, so it is drafted as a file, reviewed, executed once, and archived to setup/done/ as the record of what was migrated. (Advanced tier only; from simple tier, /axiom tier advanced comes first.)

Name the file su-<NNN>-<acr>-hierarchy-change.md (e.g. su-004-pw-hierarchy-change.md) — the unit prefix rides inside the name part, the number comes from setup/README.md as for any setup file. Beyond the standard sections it MUST contain:

  1. Unit and direction — which project or workstation, its current Levels value (read from ids.md / areas.md), and the target.
  2. Complete folder map — every sub-project/sub-workstation and every WP/WA, old path → new path, with the numbers each receives. No "and so on": the map is the instruction, and anything missing from it does not get moved.
  3. Registry rewrite — the new ids.md / areas.md in full: Levels, the sub-unit counter, and each sub-unit's own Next WP / Next area.
  4. Task ID map (project side) — old ID → new ID for every task. Task numbers (TT) are preserved; only the prefix changes.
  5. Reference sweep — the list of files holding paths or IDs that must follow: [[wiki links]] across the vault, vault/index.md, the project CLAUDE.md Structure section, the root routing map, board.md where the Todos module runs in Notion mode.
  6. Board policy (Notion mode only) — state explicitly whether cards are repushed or left untouched. Leaving them stale is a valid choice; leaving it unstated is not.
  7. Rollback — what to restore if the migration is interrupted midway.

Execution rules for these files:

  • The "monotonic, never reused" guarantee (Task ID Protocol) is suspended for this one migration and only here. Say so in the file, so the executed copy in done/ explains why numbers moved.
  • Show the user the complete diff before writing anything. This is the one setup file whose blast radius is the whole project.
  • Move folders and rewrite references in one pass, then run /lint all — broken wiki links are the expected failure mode.

After any action

  • Append to vault/log.md, update vault/index.md if pages changed.
  • Update the Last updated date in every file changed.

Notes

  • Never overwrite a filled soul.md or brand-config.md without explicit confirmation — those are owned by /axiom.
  • "new" never implements; "run" never drafts. Keep the lifecycle clean: idea → file (new) → capability (run) → archive (done/).

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.