Lint And Style Config logo

Lint And Style Config

Community
zakariaf
lint-and-style-config

Enforces a strict analysis_options.yaml built on very_good_analysis with strict-casts/strict-raw-types, a fixed set of silence-producing bug classes promoted to error (unawaited_futures, discarded_futures, empty_catches, use_build_context_synchronously, cancel_subscriptions, close_sinks, avoid_dynamic_calls, exhaustive_cases, avoid_print), dart format as the sole whitespace authority, generated-file excludes mirrored to coverage, and line-scoped-only suppression discipline; teaches the version-pinned-include trap (a missing include silently disables all rules), the errors-only-re-ranks-vs-linter-enables mechanic, and the sealed-switch analyze-vs-compile gap. Use when editing analysis_options.yaml, adding or disabling a lint, writing an `// ignore:`/`// ignore_for_file:`, bumping the Dart SDK or very_good_analysis version, wiring riverpod_lint, or explaining why discarded_futures, use_build_context_synchronously, close_sinks, or missing_provider_scope fires.

Overview

Publisherzakariaf
RepositoryFlutter-Skills
Skill namelint-and-style-config
Stars
64
Forks
15
Bundled files
6
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 6 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by zakariaf on GitHub. Read the source before you install it.

Installation

Install the Lint And Style Config AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/zakariaf/Flutter-Skills/tree/main/skills/lint-and-style-config
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/zakariaf/Flutter-Skills.git /tmp/Flutter-Skills
mkdir -p .claude/skills
cp -r /tmp/Flutter-Skills/skills/lint-and-style-config .claude/skills/lint-and-style-config
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Lint And Style Config in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Lint And Style Config on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Lint And Style Config is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Lint & Style Config

Style is not taste; it is toolchain-enforced. dart format owns whitespace, the analyzer owns correctness, and a small set of silence-producing bug classes are promoted to build-failing errors. This skill governs the analysis_options.yaml itself and the mechanics that decide whether it actually does anything. Applies whenever you edit that file, add or silence a lint, or bump the SDK/ruleset.

Read the reference for the task at hand:

  • references/rule-catalog.md — every promoted rule and its bug class, the deliberate downgrades, the contested public_member_api_docs call, and the complexity-limit table.
  • references/config-mechanics.md — the include-pin trap, errors:-re-ranks vs linter:-enables, sealed-switch analyze-vs-compile, coverage mirroring, the riverpod_lint plugin, and suppression discipline.

Copy examples/analysis_options.yaml into the package root as the starting config. Run scripts/verify-include-pin.sh and scripts/lint-gates.sh before a PR.

Non-negotiable rules

  1. dart format is the only whitespace authority. Never hand-format, never argue with it, never // dart format off except around a hand-laid table with a justifying comment. CI runs dart format --set-exit-if-changed ..
  2. Build on very_good_analysis, pinned to the version-numbered include. Never the bare analysis_options.yaml — a pub upgrade must not silently change what counts as an error. VGA already sets strict-casts, strict-inference, strict-raw-types; do not restate them in a language: block (redundant config drifts and contradicts).
  3. An info is a failure. CI runs dart analyze --fatal-infos --fatal-warnings. Unaddressed infos accumulate into noise that hides real issues.
  4. Promote only silence-producing bug classes to error — dropped Futures, swallowed failure, dead-context-after-await, resource leaks, dynamic-boundary type holes, non-exhaustive enum switches, print. Not style preferences.
  5. A promotion under errors: only works if the base ruleset already enables the rule. errors: re-ranks existing diagnostics; it cannot turn on an off rule. To enable a rule VGA omits, add it under linter: rules: first.
  6. Never mix list and map form in linter: rules:. - rule items and rule: true entries in one block is a parse error — a broken analyzer, which is a green build. To disable a base rule, use errors: <rule>: ignore.
  7. CI must build, not merely analyze. Sealed-switch exhaustiveness is a compile error, not a lint; an errors: override cannot be trusted to silence it, and any code path that reaches dart compile still fails there — an analyze-only gate can go green over code that will not build.
  8. Exclude generated code, and mirror the exact globs in coverage filtering. Otherwise generated files dilute the coverage number until it is meaningless.
  9. Suppressions are line-scoped only, with a same-line/above reason. Never // ignore_for_file: on a promoted rule; never any suppression on an architecture/import gate. Prefer fixing the code.
  10. Never add custom_lint. riverpod_lint 3.x runs on the first-party plugin system; its diagnostics fire on a plain flutter analyze.

The include, and the trap that silently disables everything

yaml
include: package:very_good_analysis/analysis_options.10.3.0.yaml

If that filename does not exist in the resolved package, you get one warning • include_file_not_found and analysis runs with zero rules. A build that checks nothing is green.

Bump the include filename in the same commit as the SDK/VGA bump, verify the file exists (ls ~/.pub-cache/hosted/pub.dev/very_good_analysis-*/lib/), and after any bump confirm a known violation still errors — never trust green. scripts/verify-include-pin.sh automates the existence check.

Promote the silence-bugs to error

yaml
analyzer:
  errors:
    unawaited_futures: error
    discarded_futures: error
    empty_catches: error
    avoid_catches_without_on_clauses: error
    only_throw_errors: error
    throw_in_finally: error
    use_build_context_synchronously: error
    cancel_subscriptions: error
    close_sinks: error            # VGA ships this at `ignore`; re-promoted
    avoid_dynamic_calls: error
    always_declare_return_types: error
    cast_nullable_to_non_nullable: error
    exhaustive_cases: error
    avoid_print: error
    avoid_slow_async_io: error

Each turns an info-level squiggle a solo dev scrolls past into a build failure. See references/rule-catalog.md for the per-rule bug class and the rules held deliberately at warning/ignore (deprecated_member_use, invalid_annotation_target, lines_longer_than_80_chars).

errors: re-ranks; linter: rules: enables

yaml
# Promoting works ONLY because VGA already enables these rules (close_sinks even
# at severity `ignore` counts as enabled — the override resurrects it).
analyzer:
  errors:
    close_sinks: error

# To enable a rule VGA does NOT ship, add it here FIRST, then optionally promote.
linter:
  rules:
    some_rule_vga_omits: true

Verify with a real violation — never assume a promotion took. The shipped config needs no linter: rules: block because every rule it promotes is VGA-enabled.

Sealed exhaustiveness is a compile error, not a lint

dart
sealed class SaveResult {}
class Saved extends SaveResult {}
class SaveFailed extends SaveResult {}

// Missing a case is `non_exhaustive_switch_statement` — a COMPILE error.
// (A switch *expression* raises `non_exhaustive_switch_expression` instead.)
String describe(SaveResult r) {
  switch (r) {
    case Saved():
      return 'saved';
    case SaveFailed():
      return 'failed';
  }
}

exhaustive_cases only covers enum / static-const sets. Model failure and mode types as sealed so a new case breaks the build at every call site instead of falling through a default into silence. An errors: override cannot be trusted to suppress an exhaustiveness diagnostic, and even where dart analyze is coaxed green dart compile still fails — so CI must build, never merely analyze. See dart3-idioms-and-coding-standards.

Excludes, mirrored to coverage

yaml
analyzer:
  exclude:
    - "**/*.g.dart"
    - "**/*.freezed.dart"
    - "**/*.drift.dart"
    - "**/generated_plugin_registrant.dart"
    - "test/drift/generated/**"   # drift schema snapshots — never reformat/fix

The identical globs must also feed coverage filtering (lcov --remove or your reporter's excludes). Excludes and coverage filters that drift apart lie the number upward. See testing-strategy and codegen-and-toolchain.

The hole no lint catches

dart
// WRONG — flagged by NOTHING. The arrow returns the Future (so discarded_futures
// is satisfied), but the target is VoidCallback, so Dart drops it silently.
onPressed: () => repository.save(note),

// RIGHT — a void-returning intent method the callback holds safely.
class NoteController {
  void saveNow(Note note) => unawaited(_repo.save(note).catchError(_report));
}
// onPressed: () => controller.saveNow(note),

The mitigation is structural, never disciplinary. A green analyzer is not proof an operation ran; the highest-severity failures are invisible to every rule and need a manual pass. See examples/callback_future_hole.dart and async-safety.

Anti-patterns

  • Bare include: package:very_good_analysis/analysis_options.yaml — a pub upgrade silently changes the ruleset; a missing pinned file silently disables it. Pin and verify.
  • Restating strict-casts/strict-raw-types in a language: block — VGA sets them; the copy drifts and eventually contradicts.
  • Promoting a rule VGA does not enable — errors: cannot resurrect an off rule, so the promotion silently does nothing.
  • Mixing - rule and rule: false in one linter: rules: block — parse error, broken analyzer, green build.
  • Analyze-only CI — passes over code that fails dart compile when a sealed switch is non-exhaustive.
  • Excludes not mirrored in coverage — generated files inflate the number.
  • // ignore_for_file: on a promoted rule, or any suppression on an import gate — the next edit is unprotected; a silenced correctness gate is a blocker.
  • Adding custom_lint — legacy plugin system; riverpod_lint 3.x needs none.
  • Reformatting a drift schema snapshot — corrupts the migration-test baseline.

Definition of done

  • dart format --set-exit-if-changed . clean; dart analyze --fatal-infos --fatal-warnings passes; CI builds (not analyze-only).
  • Include is the version-pinned VGA file and resolves in the pub cache (scripts/verify-include-pin.sh passes).
  • No redundant language: block; strict modes come from VGA.
  • The silence-bug classes are promoted to error; every promoted rule is one VGA actually enables.
  • No mixed list/map form in linter: rules:.
  • Generated-file excludes present and mirrored in coverage filtering.
  • Any suppression is line-scoped with a reason, none on an architecture gate; no // ignore_for_file: on a promoted rule.
  • riverpod_lint configured with a version: key; no custom_lint.

Related skills

  • dartdoc-conventions — backs public_member_api_docs: error and the seams to document.
  • async-safety — the Future-dropped-in-a-callback hole this config cannot see.
  • error-handling-typed-results — the sealed Failure/Result spine that makes exhaustiveness a compile-time guarantee.
  • dart3-idioms-and-coding-standards — sealed types and the exhaustive-switch discipline behind rule 7.
  • naming-conventions — the role suffixes and Effective-Dart naming the analyzer does not enforce.
  • state-management-riverpod — the Riverpod patterns riverpod_lint guards.
  • codegen-and-toolchain / run-codegen — the generated files these excludes cover.
  • ci-pipeline-and-gates — where these gates and the build-not-analyze rule run.
  • testing-strategy — coverage mirroring and what a green analyzer cannot prove.
  • dependency-hygiene — the version-pinned lint include as a dependency concern.

References

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Lint And Style Config AI skill do?

Enforces a strict analysis_options.yaml built on very_good_analysis with strict-casts/strict-raw-types, a fixed set of silence-producing bug classes promoted to error (unawaited_futures, discarded_futures, empty_catches, use_build_context_synchronously, cancel_subscriptions, close_sinks, avoid_dynamic_calls, exhaustive_cases, avoid_print), dart format as the sole whitespace authority, generated-file excludes mirrored to coverage, and line-scoped-only suppression discipline; teaches the version-pinned-include trap (a missing include silently disables all rules), the errors-only-re-ranks-vs-l...

Why use Lint And Style Config on TypingMind?

Because you install it once and use it with any model. Lint And Style Config is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Lint And Style Config in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/zakariaf/Flutter-Skills/tree/main/skills/lint-and-style-config. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Lint And Style Config?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Lint And Style Config?

As many as you like. As long as a model supports skills, you can use Lint And Style Config with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Lint And Style Config AI skill free?

Yes. It is published on GitHub by zakariaf under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇