Install Anti-Slop Oxlint Rules
dmmulroy/anti-slop
Installs, updates or migrates the vendored anti-slop Oxlint plugin in a repository, keeping local rule changes and the plugin's license and provenance files.
Implement an autofix for an existing shuck-rs lint rule. An agent skill from ewhauser/shuck.
$ npx skills add ewhauser/shuck --skill implement-fix -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install ewhauser/shuck implement-fix --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/ewhauser/shuck.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.claude/skills/implement-fix .claude/skills/implement-fix && rm -rf skills-srcUse ~/.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/
Install the "implement-fix" agent skill from https://github.com/ewhauser/shuck/tree/main/.claude/skills/implement-fix into .claude/skills/implement-fix/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "implement-fix", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/ewhauser/shuck/tree/main/.claude/skills/implement-fixType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add ewhauser/shuck --skill implement-fix -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install ewhauser/shuck implement-fix --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/ewhauser/shuck.git skills-src && mkdir -p .agents/skills && cp -r skills-src/.claude/skills/implement-fix .agents/skills/implement-fix && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "implement-fix" agent skill from https://github.com/ewhauser/shuck/tree/main/.claude/skills/implement-fix into .agents/skills/implement-fix/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "implement-fix", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add ewhauser/shuck --skill implement-fix -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install ewhauser/shuck implement-fix --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/ewhauser/shuck.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/.claude/skills/implement-fix .cursor/skills/implement-fix && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "implement-fix" agent skill from https://github.com/ewhauser/shuck/tree/main/.claude/skills/implement-fix into .cursor/skills/implement-fix/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "implement-fix", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/ewhauser/shuck.git --path .claude/skills/implement-fix--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add ewhauser/shuck --skill implement-fix -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install ewhauser/shuck implement-fix --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/ewhauser/shuck.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/.claude/skills/implement-fix .gemini/skills/implement-fix && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "implement-fix" agent skill from https://github.com/ewhauser/shuck/tree/main/.claude/skills/implement-fix into .gemini/skills/implement-fix/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "implement-fix", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install ewhauser/shuck implement-fixInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add ewhauser/shuck --skill implement-fix -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/ewhauser/shuck.git skills-src && mkdir -p .github/skills && cp -r skills-src/.claude/skills/implement-fix .github/skills/implement-fix && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "implement-fix" agent skill from https://github.com/ewhauser/shuck/tree/main/.claude/skills/implement-fix into .github/skills/implement-fix/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "implement-fix", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add ewhauser/shuck --skill implement-fix -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install ewhauser/shuck implement-fix --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/ewhauser/shuck.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/.claude/skills/implement-fix .opencode/skills/implement-fix && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "implement-fix" agent skill from https://github.com/ewhauser/shuck/tree/main/.claude/skills/implement-fix into .opencode/skills/implement-fix/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "implement-fix", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
implement-fixImplement an autofix for an existing shuck-rs lint rule. An agent skill from ewhauser/shuck.
Implement Fix is an agent skill from ewhauser/shuck. Implement an autofix for an existing shuck-rs lint rule. Use this skill whenever the user asks to add a fix to a rule (e.g., "add a fix for S074", "make C001 fixable", "implement the autofix for X023", "add safe fix for Y010"). This is about wiring an existing diagnostic up to produce edits — not about implementing new rules from scratch (that's the implement-rule skill) or fixing rule bugs (that's the fix-rule skill).
Its SKILL.md is about 3.1k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.
It sits in Development, covering Linting and formatting. The repository describes itself as: A lightning fast shell linter/formatter/LSP server with zsh support. The licence is MIT.
10 steps, taken from the step headings in SKILL.md.
Read from SKILL.md and the folder at commit 904974e. It shows what the files ask for, not the result of running them.
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.
Shell commands in SKILL.md call:
cargomakeFrom the folder's file list and the shell code blocks in SKILL.md.
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Implement Fix loads about 3.1k tokens when it runs. Until then it costs about 109 tokens; SKILL.md has 1,374 words of instructions outside code blocks.
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.
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.
The full file from ewhauser/shuck at commit 904974e, republished under its MIT licence (© ewhauser). 1,374 words, ~3,096 tokens.
.claude/skills/implement-fix/SKILL.md (or your agent's skills folder).This skill turns an existing rule into a fixable rule: it attaches a Fix to
each diagnostic, sets the right FixAvailability and applicability, adds tests
that prove the rewrite is correct, and runs the conformance corpus to make sure
the fix doesn't shift parity numbers.
The reference implementation is S074 / AmpersandSemicolon at
crates/shuck-linter/src/rules/style/ampersand_semicolon.rs. Read it first if
anything below is unclear — it covers every moving part end-to-end.
Read these files:
docs/rules/{CODE}.yaml — the YAML metadata. It tells you whether the fix is
safe and what the edit should do.crates/shuck-linter/src/rules/{category}/{rule_name}.rs).crates/shuck-linter/CLAUDE.md — the layered architecture and the "Authoring
fixes" section. The skill summarizes it but the source is canonical.crates/shuck-linter/src/fix.rs — the shared Edit, Fix, Applicability,
FixAvailability primitives.crates/shuck-linter/src/rules/style/ampersand_semicolon.rs — the reference
implementation, including its tests.If the rule lacks a fix_description in YAML, stop and ask the user — the
metadata is the source of truth and an absent fix description means the design
work hasn't been done yet.
new_code: S074
safe_fix: true # → Applicability::Safe (use Fix::safe_edit)
fix_description: "Delete the …" # → describes what the edit must doTwo fields drive everything:
| Field | Maps to |
|---|---|
safe_fix: true | Applicability::Safe, Fix::safe_edit / Fix::safe_edits |
safe_fix: false (with a fix_description) | Applicability::Unsafe, Fix::unsafe_edit / Fix::unsafe_edits |
fix_description: "..." | The behaviour the edit must produce. Use it to pick deletion vs replacement vs insertion and to phrase the fix_title(). |
The YAML is authoritative. Do not override safe_fix based on your own
judgment — if the metadata says safe, implement safe; if it says unsafe,
implement unsafe. If you genuinely think the metadata is wrong, raise it with
the user before changing the rule.
Open the rule file. The first question, lifted from crates/shuck-linter/CLAUDE.md:
Do we already have an exact span for the token/text we want to edit?
There are three cases:
; itself). Skip to Step 4.Do not rediscover the edit span from raw source inside the rule file. Rule
files cannot scan source, walk AST, or rebuild structure — the architecture test
in src/rules/mod.rs enforces this.
When the existing facts don't expose the edit span:
crates/shuck-linter/src/facts.rs (or one of the
submodules under src/facts/). Almost always one of: CommandFact,
RedirectFact, SubstitutionFact, WordFact, PipelineFact,
SimpleTestFact, ConditionalFact, or one of the option-shape facts under
CommandOptionFacts.LinterFacts::build (or the
relevant builder), and re-export any rule-facing types from the crate root
if needed.If you find yourself wanting to add a one-off helper that only this rule will ever use, that's a yellow flag — facts are shared by design, so the data should be useful to other rules too. If it really is one-off, still build it in facts; the architecture test will reject it in the rule file.
Violation implThree changes:
use crate::{Checker, Edit, Fix, FixAvailability, Rule, Violation};
impl Violation for AmpersandSemicolon {
const FIX_AVAILABILITY: FixAvailability = FixAvailability::Always; // or Sometimes
fn rule() -> Rule { Rule::AmpersandSemicolon }
fn message(&self) -> String { "...".to_owned() }
fn fix_title(&self) -> Option<String> {
Some("remove the stray `;` after `&`".to_owned())
}
}Picking FIX_AVAILABILITY:
| Variant | Use when |
|---|---|
FixAvailability::Always | Every diagnostic the rule emits will carry a fix. |
FixAvailability::Sometimes | Only some emitted diagnostics carry a fix (e.g. fix only when a specific structural shape is present). |
FixAvailability::None | No fix exists — leave as the default (don't set the constant). |
If you set Always you must actually attach a fix to every diagnostic the
rule emits. If any path emits a bare diagnostic, downgrade to Sometimes.
Picking fix_title(): a short imperative phrase describing the edit, in the
voice of "what shuck will do for you" (e.g. "remove the stray ; after &"). It
should not duplicate the diagnostic message. The YAML fix_description is a
useful starting point but tighten it for terminal output.
Replace the report* call with a report_diagnostic* call that builds a
Diagnostic and attaches a Fix:
pub fn ampersand_semicolon(checker: &mut Checker) {
let spans = checker.facts().background_semicolon_spans().to_vec();
for span in spans {
checker.report_diagnostic_dedup(
crate::Diagnostic::new(AmpersandSemicolon, span)
.with_fix(Fix::safe_edit(Edit::deletion(span))),
);
}
}Two API points to know:
Checker::report_diagnostic(diagnostic) and report_diagnostic_dedup(diagnostic)
accept a fully built Diagnostic. Use these (not report / report_dedup)
whenever you want to attach a fix.Diagnostic::new(violation, span).with_fix(fix) is the standard builder.
Prefer this over reaching into Diagnostic fields directly.Edit primitives (from src/fix.rs):
| Primitive | Use for |
|---|---|
Edit::deletion(span) | Remove a span outright (e.g. a stray ;). |
Edit::replacement(content, span) | Replace a span with new text. |
Edit::insertion(offset, content) | Insert text at an offset (zero-width). |
Edit::deletion_at(start, end) / replacement_at(start, end, content) | Same, but with raw offsets when you have them. Avoid when a span is available. |
Fix primitives:
| Primitive | Use for |
|---|---|
Fix::safe_edit(edit) / Fix::safe_edits([edit, ...]) | When safe_fix: true in YAML. |
Fix::unsafe_edit(edit) / Fix::unsafe_edits([edit, ...]) | When safe_fix: false in YAML. |
A single fix can carry multiple edits — keep them tightly related and
non-overlapping. The shared fixer in src/fix.rs handles deconflicting and
ordering across diagnostics; you do not.
For FixAvailability::Sometimes: branch in the rule body — attach the fix
only on the path where it's safe to apply, and emit a bare Diagnostic::new(...)
on the other paths.
Keep the existing positive/negative diagnostic tests. Add three more:
#[test]
fn applies_safe_fix_to_background_semicolons() {
let source = "#!/bin/sh\necho x &;\necho y & ;\n";
let result = test_snippet_with_fix(
source,
&LinterSettings::for_rule(Rule::AmpersandSemicolon),
Applicability::Safe,
);
assert_eq!(result.fixes_applied, 2);
assert_eq!(result.fixed_source, "#!/bin/sh\necho x &\necho y & \n");
assert!(result.fixed_diagnostics.is_empty());
}fixed_diagnostics must be empty — if the fix is correct, re-linting the fixed
source should not surface the same diagnostic again. If it does, the edit is
wrong.
For unsafe fixes, pass Applicability::Unsafe; the helper will then exercise
the unsafe edits.
If the rule has cases that look like the violation but should not fix (S074's
case-arm &;;/& ;; terminators), add a test that asserts fixes_applied == 0
and the source is unchanged. This protects against the fix being too greedy.
assert_diagnostics_diff!#[test]
fn snapshots_safe_fix_output_for_fixture() -> anyhow::Result<()> {
let result = test_path_with_fix(
Path::new("style").join("S074.sh").as_path(),
&LinterSettings::for_rule(Rule::AmpersandSemicolon),
Applicability::Safe,
)?;
assert_diagnostics_diff!("S074_fix_S074.sh", result);
Ok(())
}Imports needed:
use std::path::Path;
use crate::test::{test_path_with_fix, test_snippet, test_snippet_with_fix};
use crate::{Applicability, LinterSettings, Rule, assert_diagnostics_diff};crates/shuck-linter/resources/test/fixtures/{category}/{CODE}.sh should cover:
Don't widen the fixture to cover every imaginable edge case — keep it the same shape as the existing fixture and only add what the fix needs.
cargo test -p shuck-linterThe new assert_diagnostics_diff! snapshot will fail the first time. Inspect
the rendered diff in the failure output — it shows the original source, applied
edit count, unified diff, fixed source, and any remaining diagnostics. If it
matches the YAML's fix_description, accept it:
cargo insta accept --workspaceThen run the full workspace:
cargo test --workspaceIf the architecture test in src/rules/mod.rs fails, you've reached for an
AST/traversal token in a rule file — back out and push the work into facts
(Step 3).
Fixes can shift parity numbers (the diagnostic the corpus expected at a given line may be gone after the rewrite, or a previously-suppressed line may now re-trigger). Always run:
make test-large-corpus SHUCK_LARGE_CORPUS_RULES={CODE}This uses the nix-pinned ShellCheck and limits the comparison to the rule you just made fixable. If the deltas change, decide whether the new behaviour is correct (fix is doing its job and corpus expectations need updating) or whether the fix is too aggressive (back to Step 5/6).
If you can't easily tell, run the targeted comparison again with
SHUCK_LARGE_CORPUS_KEEP_GOING=1 to see every affected fixture.
Report:
FIX_AVAILABILITY chosen and why (Always vs Sometimes).Do not automatically:
LinterSettings::default_rules() — that's a separate
decision the user will make.crates/shuck/tests/ — the S074 commit added
those because it was the first fixable rule. Skip unless the user explicitly
asks.© ewhauser, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Just SKILL.md in .claude/skills/implement-fix of ewhauser/shuck.
Open the folder on GitHubat commit 904974e
Implement Fix 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.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Implement Fix this skillewhauser/shuck | 136 | — | ~3.1k | Automated safety check: Pass | MIT | |
| Install Anti-Slop Oxlint Rulesdmmulroy/anti-slop | 5.2k | — | ~2.2k | Automated safety check: Pass | MIT | |
| Babysit PR To Pass CIsgl-project/sglang | 37k | 2 repos | ~3k | Automated safety check: Pass | Apache-2.0 | |
| Rust Best Practicesfarm-fe/farm | 5.6k | 3 repos | ~1.1k | Automated safety check: Pass | MIT | |
| Go Pedantrychromedp/chromedp | 13k | — | ~3.7k | Automated safety check: Pass | MIT | |
| Kedro Babysitkedro-org/kedro | 11k | — | ~4k | Automated safety check: Pass | Custom licence |
dmmulroy/anti-slop
Installs, updates or migrates the vendored anti-slop Oxlint plugin in a repository, keeping local rule changes and the plugin's license and provenance files.
sgl-project/sglang
Start and persistently pursue a goal to babysit an SGLang pull request until selected GitHub Actions workflows pass on the latest PR head.
farm-fe/farm
Guide for writing idiomatic Rust code based on Apollo GraphQL's best practices handbook.
chromedp/chromedp
This skill should be used when the user is writing Go code and needs guidance on Go-specific pedantry: error wrapping with fmt.Errorf and %w, interface design (accept interfaces return structs)…
kedro-org/kedro
Run Kedro's local lint / format / type-check / tests on changed files (uses the project's pre-commit hooks, ruff, mypy, pytest, lint-imports, detect-secrets, Make targets — in the right venv), or…
woocommerce/woocommerce
Rules for writing and editing markdown in the WooCommerce repository, with the project's markdownlint settings for headings, lists and code blocks.
ewhauser/shuck
Profile shuck scripts and large-corpus fixtures, especially requests to profile a corpus script/fixture, reprofile after a shuck performance change, or produce a hotspot table from a samply profile.
ewhauser/shuck
Compare benchmark performance between two git worktrees (or the current worktree vs main).
ewhauser/shuck
Verify ShellCheck conformance for a shuck rule by running the large corpus test, analyzing deltas, and producing a structured bug document in docs/bugs/.
ewhauser/shuck
Fix a shuck lint rule that has conformance deltas against ShellCheck.
ewhauser/shuck
Implement a shuck-rs lint rule from its YAML definition in docs/rules/.
ewhauser/shuck
Write and update technical design specifications. An agent skill from ewhauser/shuck.
Categories
Implement an autofix for an existing shuck-rs lint rule. An agent skill from ewhauser/shuck. Implement Fix is an agent skill from ewhauser/shuck. Implement an autofix for an existing shuck-rs lint rule.
Implement Fix fits situations like: the user asks to add a fix to a rule (e.g; add a fix for S074; make C001 fixable; implement the autofix for X023.
Run `npx skills add ewhauser/shuck --skill implement-fix -a claude-code`. Or copy the skill folder (.claude/skills/implement-fix in ewhauser/shuck) into .claude/skills/implement-fix in your project. Claude Code loads it when a task matches its description.
Run `npx skills add ewhauser/shuck --skill implement-fix -a codex`. Or copy the skill folder (.claude/skills/implement-fix in ewhauser/shuck) into .agents/skills/implement-fix in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add ewhauser/shuck --skill implement-fix -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/implement-fix, .gemini/skills/implement-fix, .github/skills/implement-fix and .opencode/skills/implement-fix in your project.
Going by SKILL.md and its folder, Implement Fix needs the command-line tools its instructions call (cargo and make).
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.
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.
Implement Fix is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
About 3.1k tokens (SKILL.md is roughly 12k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
Skills that share tags, products or a category with Implement Fix: Install Anti-Slop Oxlint Rules (dmmulroy/anti-slop, 5.2k stars), Babysit PR To Pass CI (sgl-project/sglang, 37k stars), Rust Best Practices (farm-fe/farm, 5.6k stars) and Go Pedantry (chromedp/chromedp, 13k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
ewhauser (a GitHub user) maintains it in ewhauser/shuck, which has 136 GitHub stars. The repository holds 8 skills in this directory. The repository was last updated on October 5, 2026.
Source: ewhauser/shuck on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.