agentleFS
Sign inStart free
jean1880jean1880/nuvek-agent-config/skills/polyglot-architect/SKILL.md

polyglot-architect

Skill0 starsChanged 17 days agoMITView on GitHub

Multi-language scaffolding and standards: Rust, TypeScript/Vue, Python, shell, C#/.NET, C++ (start a new project, code standards, build setup, which gate and reviewer apply). For hands-on Rust use rust-craftsman.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 1 is distinctive, and 1 is 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.

  • 1

    Very specific

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

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

    Search with rg and fd, and locate the project root before acting rather than assuming paths.

    line 21
  2. Distinctive:

    If you run a local coder model, delegate short, simple code challenges, single-function drafts and unit-test authoring to it to conserve …

    line 24

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

Polyglot Architect

Triggers & Scope: Use for multi-language development standards, scaffolding and review loops across Rust, TypeScript/Vue, Python, shell, C#/.NET and C++: project structure, tooling, idioms, and which gate and reviewer apply. For hands-on Rust use rust-craftsman.

<!-- customise: set the default language per kind of project (tool, service, UI, glue) for your team. -->

LANGS: [Rust → rust-craftsman | TypeScript/Vue 3 + Tailwind v4 | Shell glue | Python → references/python.md | C#/.NET → references/dotnet.md | C++]

CHOOSING A LANGUAGE

  • A new reusable tool, service, CLI, hook or MCP server defaults to Rust.
  • Shell is thin glue only: service wrappers, hooks, orchestration and globbing.
  • A web UI is Vue 3 + Tailwind v4 on the project's shared tokens and components. A gap in the shared library is fixed in the library, not worked around in the app.
  • Python only where a project already uses it. The standards are in references/python.md.

PROJECT HYGIENE

  • Detect the project from its manifest (Cargo.toml, package.json, CMakeLists.txt, pyproject.toml) and follow its local config (rustfmt.toml, eslint, ruff). Read its AGENTS.md (or a legacy CLAUDE.md) first.
  • Search with rg and fd, and locate the project root before acting rather than assuming paths.
  • Endpoints and secrets come from the environment at runtime, never compiled in or committed.
  • A public repository never names a private deployment: no hosts, addresses, internal paths or personal identifiers.
  • If you run a local coder model, delegate short, simple code challenges, single-function drafts and unit-test authoring to it to conserve cloud tokens and session context.

REVIEW LOOP (not done without it)

  • GATE, always:
    • Rust: cargo fmt --check, cargo clippy --all-targets -- -D warnings, cargo test, or the repository's own gate script.
    • UI: the project's build, typecheck, lint and test scripts.
    • Shell: shellcheck -x.
    • Python: see the reference file.
    • C#/.NET: dotnet build then dotnet test from the repository root (details in references/dotnet.md). A red gate means not done.
  • REVIEW, as a separate pass from writing: Rust → rust-reviewer; a shared UI library → shared-library-review; shell → script-reviewer. Reviewers never edit; the author fixes and re-reviews by messaging the same agent. For heterogeneous Maker-Checker review, invoke a harness other than the author's, rotating per the delegation skill. Say which review ran. An unstated review cannot be told apart from none.
  • Scale it to the change: a logic change gets the full loop; a comment or doc-only edit gets the gate and a note.

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.

Check your setup, free