agentleFS
muslewskimuslewski/status-herald/.claude/skills/atlas-seed/SKILL.md

atlas-seed

Skill2 starsChanged 4 months agoMITView on GitHub

Use when an Atlas vault is empty or near-empty (0–2 zones after atlas init) — bootstrap 4–8 seeded zone cards from a real codebase analysis. Also use for "seed the mind", "initial zones", "first map pass", "bootstrap atlas", or atlas-fleet seed-empty.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 1

    1. muslewski/token-oracle★ 2100% same

    Edited · 0

    None found.

    Wrote their own version · 0

    None found.

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository7 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 1 is distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

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

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

    Never promote to active or stamp a SHA in this skill.

    line 19
  2. Very specific:

    After atlas-fleet ensure-all / atlas init on a new repo

    line 26
  3. Very specific:

    Do not run as a substitute for atlas-recollection after product work — that skill updates touched zones.

    line 29
  4. Distinctive:

    Prefer coherent ownership over fine-grained files:

    line 44
  5. Very specific:

    Prefer directory-level owns.globs (src/lib/) over listing every file.

    line 61
  6. Very specific:

    Every glob MUST match ≥1 git-tracked file (git ls-files '<glob>' or equivalent).

    line 62
  7. Very specific:

    Use the vault template (templates/zone.md) as base.

    line 69

≈ “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

Atlas seed — initial zone bootstrap from code

Goal: turn a scaffolded vault into a usable map after atlas init, without lying about verification.

This is the first pass, not recollection. Cards stay honest:

FieldRequired value
statusseeded
verifiedAtunverified
typezone

Never promote to active or stamp a SHA in this skill. Only a later human (or explicit atlas stamp after review) does that.

When to run

  • Vault exists (map/zones/ empty or only stubs)
  • atlas status shows 0 zones (or 1–2 toy cards)
  • After atlas-fleet ensure-all / atlas init on a new repo
  • Re-seed only if the old seed was wrong and still seeded / never verified

Do not run as a substitute for atlas-recollection after product work — that skill updates touched zones; this one creates the first partition.

Inputs to read (order)

  1. Repo README.md / AGENTS.md / package.json (or Cargo.toml / pyproject) — thesis
  2. Top-level tree (ls, not whole monorepo dump)
  3. Entry surfaces: bin/, src/, lib/, packages/, app/, cli/
  4. Tests / CI if present — good enforcedBy candidates later
  5. Existing vault map/overview.md if any

Skip: node_modules, dist, .git, huge generated assets.

How to partition (4–8 zones)

Prefer coherent ownership over fine-grained files:

Typical zoneOwns (examples)
core / enginemain lib, domain logic
cli / binbinaries, argument surface
api / serverHTTP / RPC entry
ui / webfrontend app
config / installconfig schema, installers
adapters / integrationspluggable backends
tests / cionly if tests are a product surface (e.g. contract suite)
docs / sitemarketing site in-repo only if substantial

Rules:

  • One concept per zone — not "misc" or "utils" dumps.
  • 4–8 cards for a normal repo; tiny repos may have 3; monorepos up to ~10.
  • Prefer directory-level owns.globs (src/lib/**) over listing every file.
  • Every glob MUST match ≥1 git-tracked file (git ls-files '<glob>' or equivalent). Empty globs fail atlas check.
  • No-git workspaces: still use real paths that exist on disk; note in overview that check may be limited until git init.

Zone card shape

Use the vault template (templates/zone.md) as base. Frontmatter YAML must parse under atlas's subset (flow-style arrays for nested lists):

---
type: zone
summary: "Distinctive 1–3 sentence thesis with real nouns (what + where)."
tags: [domain-tag]
status: seeded
created: YYYY-MM-DD
updated: YYYY-MM-DD
verifiedAt: unverified
owns:
  routes: []
  testids: []
  globs:
    - "src/foo/**"
    - "bin/bar.mjs"
  tools: []
depends: []
invariants: []
skills: []
related: []
sources: []
---

## What this is

Short prose. Distinctive terms for retrieval (see writing-for-retrieval).

## Anchors

Why these globs / what boundary they draw.

## Invariants

Only load-bearing rules you can point at. Prefer empty `invariants: []` over
invented rules. If you claim a rule, use flow-style enforcedBy:

  - rule: "…"
    enforcedBy: ["path/to/file.ts"]

## Lineage

README / design notes / "inferred from tree on YYYY-MM-DD seed pass".

Also follow writing-for-retrieval: summary carries distinctive terms; ## sections = one concept each.

Also write / update

  1. map/overview.md — one-paragraph product thesis + list of seeded zones
  2. atlas build — regenerate map/index.md (never hand-edit index)
  3. Optional: one tech-debt/ note if the tree has an obvious systemic gap
  4. Optional: one map/decisions/ only if the README already records a real ADR (do not invent ADRs)

Forbidden

  • status: active or commit SHA in verifiedAt without review
  • Copying personal / life content into any vault (esp. personal-knowledge-base — meta zones about repo structure only)
  • Claiming package dependencies between fleets that are only "same stack"
  • Wiping existing active zones — this skill is additive for empty/scaffold
  • Silent blanket rewrites of human-edited cards

Done criteria

atlas status    # shows N zones (N ≥ 3), all still seeded/unverified in cards
atlas build     # index lists new zones
atlas check     # no hard parse errors; ownership OK if globs match tracked files

Report: zone slug list + one line each purpose. Cards remain seeded until someone runs recollection + stamp.

Related skills

  • writing-for-retrieval — summary / section craft while authoring
  • atlas-nav — after seed, orientation uses the map
  • atlas-recollection — later sessions update zones after real work
  • atlas-adopt — brownfield vaults that already have notes, not empty init

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.