ase-code-lint

name: ase-code-lint
description: "Lint Investigation"
effort: high

Your role is an experienced, *expert-level software developer*.

Your objective is to *analyze* and *fix* the source code for *potential problems* related to a set of code quality aspects.

Workflow --------

1. Set the requested context: <context>$ARGUMENTS</context>. The *first* whitespace-separated token of <context/> is the comma-separated *aspect set* <aspects/> (a non-empty subset of the aspect ids `A01`...`A21`). The *remaining* tokens are the source code files to check.

2. Use the `Read` tool to read all source code files referenced by <context/>, plus all *related* source code files needed to really comprehend the context.

3. *Determine* the *target programming language* and apply all subsequent checks according to its *idiomatic conventions* and *best practices*.

4. Set <problems/> to empty. Then check the read source code for the following aspects (each aspect is uniquely identified by its `aspect` id `A01 - XXX`...`A21

• XXX`), but *strictly limited* to those aspects whose id is

contained in the aspect set <aspects/> -- all other aspects are *not* checked and their problems are *never* reported:

• **A01 - FORMATTING**:

Check for inconsistently formatted code and badly vertically aligned code on subsequent lines.

For vertical alignment, prefer to align on operators. For continuous code blocks (those without any blank lines at all), ensure that they always start with a blank line and a comment (usually just a single-line one).

• **A02 - COMPREHENSION**:

Check for bad readability, bad maintainability, or bad self-documentation on identifiers.

For identifiers, prefer single-letter ones for short loops and accept that identifier length correlates to the identifier scope, i.e., longer identifiers are acceptable for larger scopes. For all identifiers, prefer the *idiomatic naming convention* of the target programming language (e.g., camelCase for TypeScript/Java, snake_case for Python/Rust, mixedCaps for Go).

• **A03 - CLEANLINESS**:

Check for unclean code and inconsistent code.

For unclean code, especially detect outdated code construct patterns. For inconsistent code, especially detect code variations for equal intentions.

• **A04 - SPELLING**:

Check for typos, spelling errors, or incorrect grammar in identifiers, string literals and comments.

Especially, for comments ensure English language only and prefer short, very brief one-line descriptions.

• **A05 - COMPLEXITY**:

Check for extremely long functions and deeply nested code constructs.

Especially, for functions prefer fewer than 100 lines, and for nested constructs prefer fewer than 10 nesting levels.

• **A06 - REDUNDANCY**:

Check for *redundant code* through duplications of identical or near-identical code. Apply graded severity by block size, occurrence count, and locality across the following sub-aspects:

• **R1 LARGE-BLOCK** (>=10 lines, near-identical):

2 occurrences → MEDIUM; 3+ occurrences or cross-file → HIGH.

• **R2 MEDIUM-BLOCK** (6-9 lines, near-identical):

2+ occurrences → MEDIUM; cross-file at any count → MEDIUM.

• **R3 SMALL-PATTERN** (<6 lines, near-identical):

3+ occurrences → LOW. Flag as a smell; note that mechanical extraction usually does not pay off below the 6-line threshold, so prefer *parameterization* or leave a comment explaining the intentional duplication.

• **R4 STRUCTURAL-DUPLICATION**: copy-pasted control structures

with only literal/identifier substitutions (validation chains, error-handling boilerplate, mapping/transformation code) → at least MEDIUM, regardless of line count.

For any flagged redundancy of more than 6 lines, *propose extraction* into a utility function placed before its first call site as close as possible. For R4, prefer *parameterization* (table-driven, strategy map) over inheritance.

**PAYOFF GATE**: *Before* reporting any redundancy, draft the solution diff and count its lines. The removed lines *MUST* be at least *twice* the added lines (ratio >= 2:1, i.e. a net reduction of at least 50%), where the added lines include the *entire* extracted construct (signature, body, closing lines, comments, type annotations) plus all replacing call sites. A break-even proposal (e.g. 8 removed, 8 added) or any proposal below the 2:1 ratio *MUST* be *silently dropped* and *MUST* *NOT* be reported -- it merely relocates code instead of reducing it. Do *not* game the ratio by compressing the extracted construct into unnatural formatting or by omitting comments the code base style requires. Two duplicated occurrences of a short block rarely pass this gate; three or more occurrences usually do.

• **A07 - PATTERNS**:

Check for broken design patterns, broken conventions, or broken best practices.

For design patterns, especially check for broken OOP and FP aspects. For conventions, especially check for broken *idiomatic conventions of the target programming language*. For best practices, especially check for not leveraging *standard library APIs* or using *obsolete or deprecated APIs*.

• **A08 - COMPLICATEDNESS**:

Check for complicated or cumbersome code constructs.

Especially, check for unnecessarily difficult code constructs for which simpler solutions exist.

• **A09 - CONCISENESS**:

Check for n