agentleFS
rakaarwakyrakaarwaky/agents-arwaky/skills/architechture/aes/SKILL.md

aes

Skill0 starsChanged 8 days agoMITView on GitHub

AES docs, layers, rule fixes, and legacy migration.

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    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:

    Safety: default (no --dry-run) = autonomous.

    line 44
  2. Very specific:

    Turn progress gates: run lint-arwaky-cli report at start (baseline) and end (delta).

    line 52

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

AES Architecture

One umbrella for AES (Arwaky Engineering System) architecture work. Split by job: documents, layer scaffolding, rule fixes, migration. The Global Rules govern every job.

Global Rules

  • Split spec from status: PRD/FRD/DESIGN state intent; BACKLOG and ROADMAP state what is true now. Status words never appear in spec documents.
  • One document per reader: root documents (PRD, ROADMAP, README, AGENTS, ARCHITECTURE, CONTRIBUTING) at workspace root; FRD/BACKLOG per feature folder; DESIGN per surface folder; DATA per shared folder.
  • One fact, one home: each references/<doc>/ set owns that document's shape — rules, exemplar, template, workflow. Read it; do not restate here.
  • Verify: lint-arwaky-cli docs <workspace> passes; lint-arwaky-cli scan <layer-path> → 0 violations before moving to the next layer.
  • Layer order: migrate bottom-up (taxonomy → contract → utility → capabilities → agent → surface → root); a layer starts only when the one below scans clean.
  • Shape ownership: protocol/aggregate shape, structure, DI, wiring owned by the layer skill; migration routes; it does not restate layer shape.
  • Baseline first: lint-arwaky-cli scan . before touching anything; violation count picks strategy (< 10: full, 10–50: phased, > 50: taxonomy-first).

Jobs (labeled subsections)

1. Documents — project documentation

Scaffold and audit the ten AES project documents: PRD, ROADMAP, FRD, BACKLOG, README, AGENTS, ARCHITECTURE, CONTRIBUTING, DESIGN, DATA. Each document type has its own rules, exemplar, template, and workflow.

  • PRD/ROADMAP/FRD/BACKLOG: what/why vs how/now; spec/status split enforced.
  • README: how to use/run at workspace root.
  • AGENTS: how to work here safely at workspace root.
  • ARCHITECTURE: layer boundaries and allowed dependencies at workspace root.
  • CONTRIBUTING: how to set up and ship a change at workspace root.
  • DESIGN: visual & interaction contract per surface folder (six sections: Kind, Entry Points, Request Shape, States, Error States, Invariants, Change Checklist).
  • DATA: what the shared layer manages per shared folder.
  • Improve vs make: audit mode checks each document against its rules file and exemplar; make mode scaffolds from the template.

2. Layer — AES layer scaffolding

Scaffold AES layer files — agent orchestrators, capabilities, contracts, composition roots, surface commands/UI, taxonomy VOs, stateless utilities — across Python, Rust, TypeScript.

  • Layer chain: contract → capabilities → agent, consumed by surface/root. One layer file implements one protocol/aggregate pair.
  • Language auto-detection: resolve path, detect python/rust/typescript from files present; on mixed input, proceed with dominant language and name it in the report.
  • Routing: dispatch to reference/<layer>/<language>.md; that HOW-TO owns workflow steps.
  • Mode create: scaffold from the language HOW-TO template section.
  • Mode improve: audit against language HOW-TO rules, list findings, apply fixes.
  • Safety: default (no --dry-run) = autonomous; --dry-run = read-only report only.
  • Cross-language H2 order: Rules → Template → Section Contract → Workflow → Review → Verify (identical across all 3 languages).

3. Fix — AES rule fixes

Fix AES lint violations flagged by lint-arwaky-cli scan or lint-arwaky-cli report.

  • One fact, one home: each references/step-N-AESxxx.md carries the how-to-fix for its rule code; do not restate the same fix in two files.
  • Every reference ends with a Verify step naming the exact command.
  • Turn progress gates: run lint-arwaky-cli report at start (baseline) and end (delta). Positive delta = run not done.
  • Rule codes: AES101–AES103 (naming), AES201–AES205 (orphans), AES301–AES305 (roles/layers), AES401–AES406 (imports), AES501–AES506 (quality), AES601–AES610 (structure/documents), AES701–AES706 (project structure).

4. Migrate — legacy to AES migration

Migrate a legacy Python/Rust/TypeScript codebase into the AES 7-layer architecture.

  • Phases 0–8: audit baseline, then build each layer bottom-up, Phase 8 verify to zero.
  • Per-phase gate: lint-arwaky-cli scan <layer-dir> → 0 before next layer; language compile is fallback only.
  • Done: full scan exits 0, compile clean, tests green.

References

TopicFilePurpose
PRDreferences/prd/rules.md, exemplar.md, template.md, workflow.mdPRD document shape
ROADMAPreferences/roadmap/rules.md, exemplar.md, template.md, workflow.mdROADMAP shape
FRDreferences/frd/rules.md, exemplar.md, template.md, workflow.mdFRD shape
BACKLOGreferences/backlog/rules.md, exemplar.md, template.md, workflow.mdBACKLOG shape
READMEreferences/readme/rules.md, exemplar.md, template.md, workflow.mdREADME shape
AGENTSreferences/agents/rules.md, exemplar.md, template.md, workflow.mdAGENTS shape
ARCHITECTUREreferences/architecture/rules.md, exemplar.md, template.md, workflow.mdARCHITECTURE shape
CONTRIBUTINGreferences/contributing/rules.md, exemplar.md, template.md, workflow.mdCONTRIBUTING shape
DESIGNreferences/design/rules.md, exemplar.md, template.md, workflow.mdDESIGN shape
DATAreferences/data/rules.md, exemplar.md, template.md, workflow.mdDATA shape
Layer HOW-TOsreference/<layer>/<language>.md18 layer/language HOW-TOs (agent/capabilities/contract/root/surface/taxonomy/utility × python/rust/typescript)
Test HOW-TOsreference/test/<language>.mdTest-suite HOW-TOs for python/rust/typescript
AES fix stepsreferences/step-0-setup.md through step-41-AES706.mdFix how-to per AES rule code
Python migrationreferences/python-migration.mdPython migration playbook phases 0–8
Rust migrationreferences/rust-migration.mdRust migration playbook phases 0–8
TypeScript migrationreferences/typescript-migration.mdTypeScript migration playbook phases 0–8
Lint rulesreferences/lint-rules.mdShared lint rules and triage order

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.