Agent skill

Mas Code Standards

by AUTO-MAS-Project in AUTO-MAS-Project/AUTO-MAS

A skill your agent uses when implementing, fixing, refactoring, or reviewing non-generated AUTO-MAS code, or when preparing code-style guidance, comments, docstrings, version notes, or Conventional…

AGPL-3.0Auto-check passedDevelopment

Install Mas Code Standards

skills CLI
$ npx skills add AUTO-MAS-Project/AUTO-MAS --skill mas-code-standards -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install AUTO-MAS-Project/AUTO-MAS mas-code-standards --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/AUTO-MAS-Project/AUTO-MAS.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/mas-code-standards .claude/skills/mas-code-standards && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
mas-code-standards
GitHub stars
707
Token cost
~2.5k tokens
SKILL.md length
1,384 words
Files
3 (incl. references)
Skills in repo
15
Repo updated
First seen
Licence
AGPL-3.0

At a glance

A skill your agent uses when implementing, fixing, refactoring, or reviewing non-generated AUTO-MAS code, or when preparing code-style guidance, comments, docstrings, version notes, or Conventional…

  • Works in 3 steps: frontend/electron/services → frontend/electron/ipc → frontend/src/views/Initialization
  • Reviewing non-generated AUTO-MAS code
  • SKILL.md covers Objective, Scope, Workflow and Commit Lenses, plus 4 more sections
  • Calls python

What it does

Mas Code Standards is an agent skill from AUTO-MAS-Project/AUTO-MAS. Use when implementing, fixing, refactoring, or reviewing non-generated AUTO-MAS code, or when preparing code-style guidance, comments, docstrings, version notes, or Conventional Commit wording. Skip read-only diagnosis, explanation, exploration, and planning unless code conventions are requested.

Its SKILL.md is about 2.5k tokens, which your agent loads only when the skill is triggered. The skill folder holds 4 other files, including reference files (for example `agents/openai.yaml` and `references/style-observations.md`).

It sits in Development, covering Changelog and release notes, Technical documentation and Commit messages. The repository describes itself as: 多脚本多配置统一管理与自动化工具 | 轻松管理大量脚本并存储多个用户配置、设计自动化任务流、监看脚本日志,大幅提高自动化代理效率与稳定性!. The licence is AGPL-3.0.

When your agent uses it

  • Reviewing non-generated AUTO-MAS code
  • Preparing code-style guidance
  • Conventional Commit wording

Example prompts

  • “/mas-code-standards”

Requirements

  • Python 3

Workflow steps

3 steps, taken from the first numbered list in SKILL.md.

  1. frontend/electron/services
  2. frontend/electron/ipc
  3. frontend/src/views/Initialization

What it can do on your machine

Read from SKILL.md and the folder at commit 699de5a. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • python

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Mas Code Standards loads about 2.5k tokens when it runs, and up to ~4.9k if it reads all its reference files. Until then it costs about 79 tokens; SKILL.md has 1,384 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~79
When it runs · the whole SKILL.md, loaded when a task matches
~2.5k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~4.9k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from AUTO-MAS-Project/AUTO-MAS at commit 699de5a, republished under its AGPL-3.0 licence (© AUTO-MAS-Project). 1,384 words, ~2,485 tokens.

Download SKILL.mdSave it as .claude/skills/mas-code-standards/SKILL.md (or your agent's skills folder). This skill also uses 2 other files; get the full folder from GitHub.
name
mas-code-standards
description
Use when implementing, fixing, refactoring, or reviewing non-generated AUTO-MAS code, or when preparing code-style guidance, comments, docstrings, version notes, or Conventional Commit wording. Skip read-only diagnosis, explanation, exploration, and planning unless code conventions are requested.

MAS Code Standards

Objective

Reproduce the practical implementation style used in current dev without copying obsolete behavior or broadening scope unnecessarily.

Scope

Primary style reference samples:

  1. frontend/electron/services
  2. frontend/electron/ipc
  3. frontend/src/views/Initialization

Use these samples as style lenses only. For frontend engineering or UI decisions, prefer mas-frontend-standards and mas-frontend-ui; for Python modules, carry over the same values of explicit orchestration, compatibility-first changes, and operational logging, but always compare nearby files before applying style assumptions.

Workflow

  1. Read references/style-observations.md, with priority on commit lenses e541fa5f, 727aafb, and e5d72bdb.
  2. Sample 2 to 3 sibling files in the same module before editing.
  3. Keep the main execution path obvious; extract helpers only when they make the flow easier to follow.
  4. Match nearby naming, logging tone, comment style, and result contracts.
  5. Prefer minimal edits that blend into surrounding code rather than style-driven rewrites.
  6. If recent maintainer review comments are available for the same area, treat them as the strongest style signal.
  7. Before completing a user-visible feature or fix, add exactly one changelog fragment under changelog.d/ (<PR number or branch>.<feat|change|fix|breaking|remove|security|dev>.md; first line project: <key> from the project table in changelog.d/README.md — 14 adapters plus home / scheduler / emulator / notify / tools / settings / update / runtime, no catch-all, pick the nearest per the README; only dev fragments may omit it; then one user-facing sentence of at most 50 characters without project prefix, PR number or signature; python scripts/changelog.py add <type> <project> "<sentence>" creates it, with - as the project for dev). Before writing a fix or change fragment, decide which release introduced the feature you touch: a feature absent from the previous stable release's notes was introduced in the current X.Y.0 cycle (for 5.5.0: Runtime initialization, virtual display, Emulator 2.0, MFW, BetterGI, BAAH, ZZZ-OD, config restore, operator cultivation). Every later fix, maintenance or supplementary change to such a feature (bug fixes, turning something into a prompt, extra hints, default tweaks, layout changes, i18n wiring, cleanup) is beta-only: if the feature's entry is still in the top ## [未发布] section or its fragment is still in changelog.d/, add no fragment and label the PR skip-changelog (describe any wording change in the PR for a maintainer to apply); if it shipped in an earlier beta of the cycle, add the header line beta-only: true so the entry is dropped from the stable roll-up. Sub-features added to it later are beta-only as well: in the stable notes a new adapter is exactly one line (【bgi】新增 bgi 专项) and a new feature is its single introduction entry; fixes and changes to features that shipped in the previous stable are normal. Project keys use the short names bgi (BetterGI) and end (MaaEnd). A PR with several human authors lists them all in the header line author: a, b (the script never reads Co-authored-by). Maintainers may add highlight: true to route an entry into 「本次亮点」; treat it as part of the implementation rather than a reminder. Condense all changes of the PR into that one sentence. Never edit CHANGELOG.md, res/version.json or any version number: they are written only by the release PR.

Commit Lenses

  1. e541fa5f: small cleanup. Keep logic local, remove one-off indirection, simplify conditions without changing outcome.
  2. 727aafb: feature landing. Extend existing books, models, managers, routes, forms, and logs in parallel instead of inventing a separate architecture for the new type.
  3. e5d72bdb: dev integration. Preserve the feature branch structure, then consolidate at integration points instead of rewriting during merge.

Core Traits

  1. Use clear file sections such as // ==================== 类型定义 ====================.
  2. Keep module headers and comments short, Chinese, and purpose-driven.
  3. Prefer explicit orchestration over generic abstractions or clever helpers.
  4. Use getLogger('中文名') and short operational log lines.
  5. Allow light duplication when it keeps each step readable and local.
  6. Respect file-local formatting; do not normalize unrelated code.
  7. When adding a new script or domain type, extend config model, schema, routing, task registration, frontend types, composables, and edit views together.
  8. Prefer compatibility edits at registration points such as BOOK, union types, routing branches, and progress payloads before deeper refactors.
  9. For finite variants, prefer a small dict/registry mapping instead of many near-identical branches.
  10. Keep frequently edited task-configuration logic visible in the owning flow with a short comment instead of hiding it behind one-off helpers.
  11. Do not copy another script/domain's special-case logs, timeout exemptions, ignore lists, or workaround branches into a new module until that behavior is confirmed locally.
  12. If a new capability is used only once, default to an inline block or local helper; split into builder, loader, or extra services only after real reuse or boundary pressure appears.
  13. Trust existing validators, config containers, and task bases when they already guarantee an invariant; do not add a second layer of fallback or correction.
  14. For new Python-heavy flows, keep signatures and call sites compatible with at least basic static type checking.
  15. Prefer deleting redundant imports, waits, and wrappers over preserving "explicit" but noisy scaffolding.
  16. Use Conventional Commits for commit messages: <type>(<scope>): <subject>, with documented types such as feat, fix, docs, style, refactor, perf, test, chore, build, and ci.
  17. Choose commit scope from the touched file name when one file changes, or from the parent folder name when multiple related files change.
  18. For backend docstrings, use Google-style sections for summary, Args, Returns, and Raises when a function needs explanation.
  19. For ConfigBase subclasses, comment every ConfigItem and group config items by clear section markers before super().__init__().
Show full SKILL.md (478 more words)Show less

Comment Preservation

Do not treat cleanup as permission to strip comments. Preserve comments that explain business intent, operational steps, compatibility decisions, non-obvious invariants, maintainer context, or config-field meaning.

Only remove or rewrite a comment when it is stale, misleading, mechanically restates the next token of code, or refers to code that is being deleted in the same edit. When rewriting, keep the useful intent and make it shorter or more accurate instead of dropping it.

Protect these comments especially:

  1. ConfigItem comments and section markers in ConfigBase subclasses.
  2. Ordered workflow comments such as "第一步" and "第二步" in long service flows.
  3. Task-injection, config-swap, log-monitoring, and compatibility notes in runtime code.
  4. Frontend section comments that help scan large edit pages or distinguish type-specific logic.

Avoid

  1. Do not introduce framework-heavy abstractions or generic factories unless the surrounding module already uses them.
  2. Do not hide the main workflow inside too many helper layers.
  3. Do not switch comment or logging language inconsistently inside a file.
  4. Do not turn a small fix into a broad refactor for stylistic purity.
  5. Do not confuse "explicit" with "verbose".
  6. Do not modify OpenAPI-generated files. Ask the developer to regenerate them manually when updates are required.
  7. Do not split equivalent behavior into multiple methods or API routes when a dict selector or type field represents the difference directly.
  8. Do not leave dead support paths for future detailed/raw config when there is no complete UI-to-runtime path yet.
  9. Do not add maintainer-facing "safety" code that only repeats guarantees already enforced by base classes or validators.
  10. Do not complete a user-visible feature or fix without adding its changelog fragment under changelog.d/.
  11. Do not delete useful comments just to make a diff look cleaner.

Review Checklist

  1. The new code reads like neighboring project-standard files.
  2. Logs and comments are concise, operational, and consistent with the touched module.
  3. The main path is still easy to trace top-to-bottom.
  4. Compatibility and existing behavior were preserved unless the task explicitly changed them.
  5. The chosen style lens matches the task: cleanup, feature landing, or dev integration.
  6. Finite variant selection uses existing books/registries/dicts where that is clearer than method proliferation.
  7. Config choices have one source of truth and do not duplicate existing mode/tab selectors.
  8. New special cases were verified for this domain instead of cargo-culted from another module.
  9. New code would survive basic static type checking without relying on None or fallback branches that conflict with declared types.
  10. Commit messages and scopes follow the project convention when preparing commits.
  11. Backend comments and config-class annotations use the project style rather than ad hoc prose.
  12. Existing useful comments were preserved or updated accurately, not removed as noise.
  13. User-visible features and fixes include exactly one changelog fragment under changelog.d/; CHANGELOG.md, res/version.json and version numbers stay untouched.

© AUTO-MAS-Project, AGPL-3.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 2 other files (references) in .agents/skills/mas-code-standards of AUTO-MAS-Project/AUTO-MAS.

  • SKILL.md
  • agents/openai.yaml
  • references/style-observations.md

Open the folder on GitHubat commit 699de5a

Compare with similar skills

Mas Code Standards next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Mas Code Standards compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Mas Code Standards this skillAUTO-MAS-Project/AUTO-MAS707—~2.5kAutomated safety check: PassAGPL-3.0
Swig Conventionsswig/swig6.3k—~2.7kAutomated safety check: PassCustom licence
CommitLennartHennigs/Button2565—~562Automated safety check: PassMIT
CommitLennartHennigs/ESPRotary188—~590Automated safety check: PassMIT
Proseploxc/modbux108—~2.1kAutomated safety check: PassMIT
Google Devdocs StyleEpicenterHQ/epicenter4.8k—~2.8kAutomated safety check: PassCustom licence

Similar skills

  • SWIG source and contribution conventions: clang-format / code formatting, C/C++ comment style (quotes, widths, function header blocks), parser.y new-code rules, alphabetical ordering of makefile…

    6.3k GitHub stars~2.7k tokensUpdated yesterday
    DevelopmentAuto-check passed
  • Commit

    LennartHennigs/Button2

    Stage and commit current changes for Button2 — checks for needed CHANGELOG/README/CLAUDE.md updates, creates a branch if on master, writes a commit message, and commits

    565 GitHub stars~562 tokensUpdated 4 mo ago
    DevelopmentAuto-check passed
  • Commit

    LennartHennigs/ESPRotary

    Stage and commit current changes for ESPRotary — checks for needed CHANGELOG/README/CLAUDE.md updates, creates a branch if on master, writes a commit message, and commits

    188 GitHub stars~590 tokensUpdated 2 mo ago
    DevelopmentAuto-check passed
  • Prose

    ploxc/modbux

    Measure a sentence you just wrote against the thing it describes, then prune the block it lands in.

    108 GitHub stars~2.1k tokensUpdated yesterday
    DevelopmentAuto-check passed
  • Google Devdocs Style

    EpicenterHQ/epicenter

    Write and review developer documentation in Google Developer Documentation Style.

    4.8k GitHub stars~2.8k tokensUpdated today
    DevelopmentAuto-check passed
  • Fantasia Final Cleanup

    vishiri/fantasia-archive

    End-of-batch ship workflow for Fantasia Archive: run full yarn testbatch:verify (not dev scoped gate), fix failures, sync README/AGENTS/rules/skills from Git changes, update in-app changelog, split…

    409 GitHub stars~1.2k tokensUpdated 6 days ago
    DevelopmentAuto-check passed

More from AUTO-MAS-Project/AUTO-MAS

All 15 skills in this repo
  • Mas Game Sign

    AUTO-MAS-Project/AUTO-MAS

    Add, refactor, or review AUTO-MAS game community sign-in (game sign) code, including the provider registry in app/tools/gamesign.py, platform adapters for Skland/Miyoushe/Kuro/Taygedo, credential…

    707 GitHub stars~2.4k tokensUpdated yesterday
    Auto-check passed
  • Mas Schema Naming

    AUTO-MAS-Project/AUTO-MAS

    Define canonical naming for future backend schema domains. An agent skill from AUTO-MAS-Project/AUTO-MAS.

    707 GitHub stars~1k tokensUpdated yesterday
    Auto-check passed
  • Mas Frontend UI

    AUTO-MAS-Project/AUTO-MAS

    A skill your agent uses when working on AUTO-MAS frontend UI, Ant Design Vue components, page layout, forms, tables, modals, drawers, feedback, empty/loading/error states, drag interactions, dark…

    707 GitHub stars~3.8k tokensUpdated yesterday
    Auto-check passed
  • Mas Script Specialized Adapter

    AUTO-MAS-Project/AUTO-MAS

    Review, add, or refactor AUTO-MAS specialized script adapters by upstream architecture, including MAA, SRC, MaaEnd/MXU, General, ok-script adapters such as Okww and OkNte, multi-engine adapters such…

    707 GitHub stars~2.8k tokensUpdated yesterday
    Auto-check passed
  • Mas API Contract

    AUTO-MAS-Project/AUTO-MAS

    Define backend API contract standards for FastAPI services. An agent skill from AUTO-MAS-Project/AUTO-MAS.

    707 GitHub stars~1.9k tokensUpdated yesterday
    Auto-check passed
  • Mas Data Model

    AUTO-MAS-Project/AUTO-MAS

    Define backend data modeling standards for Python services. An agent skill from AUTO-MAS-Project/AUTO-MAS.

    707 GitHub stars~1.6k tokensUpdated yesterday
    Auto-check passed

Categories

Questions about Mas Code Standards

What does Mas Code Standards do?

A skill your agent uses when implementing, fixing, refactoring, or reviewing non-generated AUTO-MAS code, or when preparing code-style guidance, comments, docstrings, version notes, or Conventional…. Mas Code Standards is an agent skill from AUTO-MAS-Project/AUTO-MAS. Use when implementing, fixing, refactoring, or reviewing non-generated AUTO-MAS code, or when preparing code-style guidance, comments, docstrings, version notes, or Conventional Commit wording.

When should I use Mas Code Standards?

Mas Code Standards fits situations like: reviewing non-generated AUTO-MAS code; preparing code-style guidance; conventional Commit wording.

How do I install Mas Code Standards in Claude Code?

Run `npx skills add AUTO-MAS-Project/AUTO-MAS --skill mas-code-standards -a claude-code`. Or copy the skill folder (.agents/skills/mas-code-standards in AUTO-MAS-Project/AUTO-MAS) into .claude/skills/mas-code-standards in your project. Claude Code loads it when a task matches its description.

How do I install Mas Code Standards in Codex?

Run `npx skills add AUTO-MAS-Project/AUTO-MAS --skill mas-code-standards -a codex`. Or copy the skill folder (.agents/skills/mas-code-standards in AUTO-MAS-Project/AUTO-MAS) into .agents/skills/mas-code-standards in your project. Codex loads it when a task matches its description.

Can I use Mas Code Standards in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add AUTO-MAS-Project/AUTO-MAS --skill mas-code-standards -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/mas-code-standards, .gemini/skills/mas-code-standards, .github/skills/mas-code-standards and .opencode/skills/mas-code-standards in your project.

What does Mas Code Standards need to run?

Going by SKILL.md and its folder, Mas Code Standards needs the command-line tools its instructions call (python). Our summary lists: Python 3.

Does Mas Code Standards access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Mas Code Standards safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Mas Code Standards use?

Mas Code Standards is published under the AGPL-3.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Mas Code Standards use?

About 2.5k tokens (SKILL.md is roughly 9.9k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 2.4k tokens, read only when the agent opens those files.

What are the alternatives to Mas Code Standards?

Skills that share tags, products or a category with Mas Code Standards: Swig Conventions (swig/swig, 6.3k stars), Commit (LennartHennigs/Button2, 565 stars), Commit (LennartHennigs/ESPRotary, 188 stars) and Prose (ploxc/modbux, 108 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Mas Code Standards?

AUTO-MAS-Project (a GitHub organization) maintains it in AUTO-MAS-Project/AUTO-MAS, which has 707 GitHub stars. The repository holds 15 skills in this directory. The repository was last updated on October 7, 2026.

Source: AUTO-MAS-Project/AUTO-MAS on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.