# Settings > Every key, accepted value, and default of gspot.toml. Every key of `gspot.toml`, its accepted values, and its default. See [gspot.toml](/configure/gspot-toml/) to change them. The [JSON Schema](/gspot.schema.json) describes every key. Editors apply it on their own once SchemaStore lists gspot. ## Top-level keys ### `level` `"recommended"`, `"all"`. Default: `"recommended"`. Correctness, security, accessibility, type safety, routine formatting, and the project’s own declared checks apply at both levels. Level `all` adds checks and rules for word choice, architecture, naming, doc comments, declaration order, API style, and complexity. No level turns on experimental or preview lint rules. ### `allowed_words` List or a reasoned value `{ value = …, reason = "…" }`. Default: `[]`. Reason needed when: adding. Words accepted beyond the shipped spelling list; adding needs a reason. ### `tool_timeout_seconds` Whole number. Default: `600`. The time limit in seconds for every tool command, including installation and version reads. Root only. ### `test_files` List. Default: `["**/test/**","**/tests/**","**/__tests__/**","**/*.test.*","**/*.spec.*","**/test_*.py","**/*_test.py","**/conftest.py","**/Tests/**","**/*Tests.swift"]`. Test files. Linters apply their test rules to these files. ### `configurations` List. Default: `[]`. The configurations this repository uses besides the always-selected ones. `gspot init` writes the detected ones here; `gspot add` and `gspot remove` change the list. ### `removed_configurations` List. Default: `[]`. Configurations you removed, which detection does not add back; only `gspot add` and `gspot remove` write this list. ## \[\[scope]] Settings for one folder: its configurations and the tables that differ from the root. See [Monorepos and scopes](/configure/monorepos/) for how a scope inherits settings. ### `scope.path` Text. Required. The folder of the scope; no two records may name the same folder. ### `scope.configurations` List. Default: `[]`. Configurations for this folder, on top of the root ones; `gspot add --scope` writes this list. ### `scope.removed_configurations` List. Default: `[]`. Configurations removed in this folder; only `gspot add` and `gspot remove` write this list. ### `scope.test_files` List. Default: unset. More test files for this folder, added to the root list. Only the root takes these keys: `level`, `allowed_words`, `tool_timeout_seconds`, `import_rule`, `ignore`, `command_check`, `hooks`, `ci`, `rules`. ## \[format] Indentation, line width, quotes, and line endings for every formatter, with overrides for some paths. ### `format.indent_style` `"space"`, `"tab"`. Default: `"space"`. The character of one indent level: space or tab. ### `format.indent_width` Whole number. Default: `4`. Spaces per indent level. ### `format.print_width` Whole number. Default: `120`. The line width formatters wrap at. ### `format.line_ending` `"lf"`, `"crlf"`. Default: `"lf"`. The line ending every file uses: lf or crlf. ### `format.final_newline` `true` or `false`. Default: `true`. Whether every file ends with a newline. ### `format.quotes` `"single"`, `"double"`. Default: `"single"`. Single or double, for the languages where a formatter chooses. ### `format.trailing_commas` `"all"`, `"es5"`, `"none"`. Default: `"all"`. Where trailing commas go: all, es5, or none. ### `format.semicolons` `true` or `false`. Default: `true`. Whether JavaScript and TypeScript statements end with semicolons. ### \[\[format.override]] Formatting options for repository-relative paths, in order. #### `format.override.paths` List. Required. The files these formatting values apply to. #### `format.override.indent_style` `"space"`, `"tab"`. Default: unset. The character of one indent level: space or tab. #### `format.override.indent_width` Whole number. Default: unset. Spaces per indent level. #### `format.override.print_width` Whole number. Default: unset. The line width formatters wrap at. #### `format.override.line_ending` `"lf"`, `"crlf"`. Default: unset. The line ending every file uses: lf or crlf. #### `format.override.final_newline` `true` or `false`. Default: unset. Whether every file ends with a newline. #### `format.override.quotes` `"single"`, `"double"`. Default: unset. Single or double, for the languages where a formatter chooses. #### `format.override.trailing_commas` `"all"`, `"es5"`, `"none"`. Default: unset. Where trailing commas go: all, es5, or none. #### `format.override.semicolons` `true` or `false`. Default: unset. Whether JavaScript and TypeScript statements end with semicolons. ## \[limits] Size and complexity limits for files and functions, with values per language. Change a limit for the repository, or use `limits..` for one language. A value that loosens a check needs a reason. ```shell gspot set limits.block_nesting 4 --reason "The parser is one state machine." ``` ```toml configurations = ["bash"] level = "all" [limits] block_nesting = { value = 4, reason = "The parser is one state machine." } ``` For one language, set `limits.bash.block_nesting` instead. Its recommended default is `3`. For a Bash function over this nesting limit, `bash/function-size` reports: ```text main has 4 levels of nesting, over the limit of 3. ``` The raised limit accepts that depth. To inspect the current value and source, run `gspot explain limits.block_nesting`. ### `limits.file_lines` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `300`. Reason needed when: raising. Language defaults: sql `400`, bash `140`. The most lines one file may have, not counting blank lines and comment-only lines. ### `limits.function_lines` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `60`. Reason needed when: raising. Language defaults: bash `40`. The most lines one function may have. ### `limits.function_parameters` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `7`. Reason needed when: raising. The most parameters one function may take. ### `limits.function_statements` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `30`. Reason needed when: raising. The most statements one function may have. ### `limits.block_nesting` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `3`. Reason needed when: raising. Language defaults: bash `3`. How deep blocks may nest inside one function. ### `limits.callback_nesting` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `3`. Reason needed when: raising. How deep callbacks may nest. ### `limits.cognitive_complexity` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `8`. Reason needed when: raising. The highest sonarjs cognitive score one function may have; the score grows as the function gets harder to follow. ### `limits.cyclomatic_complexity` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `8`. Reason needed when: raising. The most paths that may run through one function. ### `limits.function_branches` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `8`. Reason needed when: raising. Language defaults: python `8`, bash `8`. The most branches one function may have. ### `limits.function_returns` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `4`. Reason needed when: raising. Language defaults: python `4`. The most return statements one function may have. ### `limits.function_assignments` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `8`. Reason needed when: raising. Language defaults: bash `8`. The most variable assignments one shell function may make. ### `limits.closure_lines` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `60`. Reason needed when: raising. Language defaults: swift `60`. The most lines one Swift closure may have. ### `limits.type_lines` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `300`. Reason needed when: raising. Language defaults: swift `300`. The most lines one Swift type body may have. ### `limits.index_exports` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `20`. Reason needed when: raising. The most re-exports one index file may hold. ### `limits.trivial_function_statements` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `2`. Reason needed when: lowering. Report functions with this many statements or fewer; any whole number from 1 up works, a higher number reports more functions, and limits.`` sets it for one language. ### `limits.shared_prefix_files` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `2`. Reason needed when: raising. The number of siblings sharing a name prefix that makes a finding. ### `limits.` The keys of `[limits]` for one language: `typescript`, `javascript`, `python`, `bash`, `swift`, `sql`. A key you leave out takes the value of `[limits]`. ## \[naming] Banned words, case, and length for the names of files, folders, and identifiers, with values per language. Allow an exact name only when an external contract requires it. Every shipped banned naming group applies at level `all`. The `roles` group rejects vague job labels such as `helper` and `manager`; `defensive` rejects qualifiers such as `ensure` and `maybe`; `marketing` rejects claims such as `advanced` and `enhanced`. `naming.banned_words` adds project words. For `export const helperCommand = 1;`, `naming/identifiers` reports: ```text javascript variable "helperCommand": "helper" is banned (roles group). ``` This complete example accepts that exact variable name in one file: ```toml configurations = ["javascript", "naming"] level = "all" [[naming.override]] paths = ["helper.js"] categories = ["variables"] allowed_names = ["helperCommand"] reason = "An external API requires this exact name." ``` Write override records by hand, then run `gspot apply`. The override removes this `banned-word` finding; other names keep their checks. To change the word limit, run: ```shell gspot set naming.max_words 5 --reason "The public API uses five words." ``` The `words` finding reports identifiers over the default of 4 words. The command permits 5 words. ### `naming.max_characters` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `35`. Reason needed when: raising. Language defaults: swift `40`, sql `55`. The most characters one identifier may have. ### `naming.max_words` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `4`. Reason needed when: raising. Language defaults: swift `5`, sql `7`. The most words one identifier may have. ### `naming.cases` List or a reasoned value `{ value = …, reason = "…" }`. Default: `[]`. Reason needed when: adding. The case patterns that one identifier category accepts: camel, pascal, kebab, snake, upper-snake, pascal-extension (String+Format.swift), or timestamp-snake (20260103120000\_create\_users.sql). Ruff checks the case of Python classes, functions, and parameters, and SwiftLint checks Swift types, variables, parameters, and enum cases; this setting covers the rest. ### `naming.banned_words` List. Default: `[]`. Words this repository bans in identifiers, beyond the shipped groups. ### \[\[naming.override]] Naming rules for repository-relative paths, in order; each record has its reason. #### `naming.override.paths` List. Default: `["**"]`. Where the override applies. #### `naming.override.languages` List. Default: unset. The languages it applies to. #### `naming.override.categories` List. Default: unset. The kinds of name it applies to. Unknown kinds are refused. #### `naming.override.ignored_prefix` Text. Default: unset. A prefix not counted as a word, such as ^test\_. #### `naming.override.allow_digits` `true` or `false`. Default: `false`. Accept digits in names. #### `naming.override.allow_repeated_words` `true` or `false`. Default: `false`. Accept a word twice in one name. #### `naming.override.allowed_names` List. Default: unset. Names accepted despite the naming rules. #### `naming.override.cases` List. Default: unset. The cases accepted here. #### `naming.override.reason` Text. Required. Why these names may break the rules. ### `naming.` The keys of `[naming]` for one language: `typescript`, `javascript`, `python`, `swift`, `bash`, `sql`, `css`, `html`. A key you leave out takes the value of `[naming]`. `naming..` sets the same keys for one category of names, and a key you leave out takes the value of `naming.`. The categories of each language: | Language | Categories | | ------------ | --------------------------------------------------------------------------------- | | `typescript` | `files`, `folders`, `types`, `functions`, `parameters`, `variables`, `properties` | | `javascript` | `files`, `folders`, `types`, `functions`, `parameters`, `variables`, `properties` | | `python` | `files`, `folders`, `types`, `functions`, `parameters`, `variables`, `properties` | | `swift` | `files`, `folders`, `types`, `functions`, `parameters`, `variables`, `properties` | | `bash` | `files`, `folders`, `functions`, `variables` | | `sql` | `files`, `folders`, `functions`, `parameters` | | `css` | `files`, `folders` | | `html` | `files`, `folders` | ## \[cloudflare] The file that holds the generated Cloudflare binding types, and its interface name. ### `cloudflare.types_file` Text. Default: `"worker-configuration.d.ts"`. Tracked Worker environment types file, relative to the repository root. ### `cloudflare.types_interface` Text. Default: `"Env"`. Environment interface name passed to wrangler types; OpenNext uses CloudflareEnv. ## \[coverage] The lowest test coverage that the test checks accept. ### `coverage.lines` Number or a reasoned value `{ value = …, reason = "…" }`. Default: recommended: `0`; all: `80`. Reason needed when: lowering. The minimum percentage of source lines covered by tests. ### `coverage.branches` Number or a reasoned value `{ value = …, reason = "…" }`. Default: recommended: `0`; all: `80`. Reason needed when: lowering. The minimum percentage of source branches covered by tests. ### `coverage.functions` Number or a reasoned value `{ value = …, reason = "…" }`. Default: recommended: `0`; all: `80`. Reason needed when: lowering. The minimum percentage of source functions covered by tests. ### `coverage.statements` Number or a reasoned value `{ value = …, reason = "…" }`. Default: recommended: `0`; all: `80`. Reason needed when: lowering. The minimum percentage of source statements covered by tests. ### \[\[coverage.swift\_target]] Line coverage floors for named Swift targets. #### `coverage.swift_target.name` Text. Required. A Swift target named in the coverage report. #### `coverage.swift_target.lines` Number. Required. The least share of lines that the target must cover, in place of `coverage.lines`. #### `coverage.swift_target.reason` Text. Default: unset. Why this target may cover less. ## \[dependencies] How many days a new release waits before an install picks it up, and the registries that lockfiles may use. ### `dependencies.min_release_age_days` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `7`. Reason needed when: lowering. How many days a new release waits before gspot installs it as one of its tools in .gspot. ### `dependencies.allowed_registries` List or a reasoned value `{ value = …, reason = "…" }`. Default: `["registry.npmjs.org","registry.yarnpkg.com"]`. Reason needed when: adding. The hosts a lockfile may resolve packages from. ### `dependencies.bun_scanner` Text. Default: unset. The npm package that Bun runs as its install security scanner. Without it, Bun runs no scanner. ## \[docs] Headings that documentation may not use, and URLs that the link check accepts. ### `docs.banned_headings` List. Default: `[]`. Headings this repository bans beyond the shipped list. ### `docs.allowed_urls` List or a reasoned value `{ value = …, reason = "…" }`. Default: `[]`. Reason needed when: adding. Allowed URL patterns; additions need a reason for the whole list. ## \[drizzle] The names of Drizzle clients in your code. ### `drizzle.client_names` List. Default: `["db"]`. The Drizzle client identifiers checked for update and delete operations without a where clause. ## \[javascript] The script files, which may print to the console and exit. ### `javascript.script_files` List. Default: `["scripts/**","**/*.config.{js,mjs,cjs,ts}"]`. Script files, which may print to the console and call process.exit. ## \[trpc] The files that declare tRPC procedures, and the server files that other files import only as types. ### `trpc.router_files` List. Default: `["**/routers/**","**/*router*.ts","**/trpc/**"]`. The files that declare tRPC procedures. ### `trpc.server_files` List. Default: `[]`. The server files that other files of a tRPC scope may import only as types; empty turns the rule off. ## \[zustand] The files that may create a Zustand store. ### `zustand.store_files` List. Default: `["**/store.*","**/stores/*","**/store/*","**/*.store.*"]`. The files that may create a store. ## \[licenses] The licenses that dependencies may have, and single packages accepted with a reason. ### `licenses.allowed` List. Default: `[]`. The SPDX license IDs a dependency may report. ### \[\[licenses.allowed\_package]] Packages accepted under a license outside the list: name, exact version, and reason. #### `licenses.allowed_package.name` Text. Required. The package name, such as `lightningcss` or `@fontsource-variable/geist`. #### `licenses.allowed_package.version` Text. Required. The exact installed version; gspot reports a record that matches no installed package. #### `licenses.allowed_package.reason` Text. Required. Why you accept its license; name the license here. ## \[openapi] The committed OpenAPI document, and the command that generates it. ### `openapi.document` Text. Default: unset. The committed OpenAPI document, relative to the repository root, which init fills in when it finds one; without it, the OpenAPI checks do not run. ### `openapi.generate_command` List. Default: unset. The command that writes the OpenAPI document from the code, run in the project scope folder. ## \[postgres] The migrations folder, the schemas that clients read, and the migrations that may not change. ### `postgres.client_schemas` List. Default: `[]`. The schemas clients reach, where every table needs row level security. ### `postgres.frozen_through` Text. Default: `"none"`. The last migration version that has run everywhere (none, all, or a version); Squawk skips those migrations, and they may not change. ### `postgres.migrations_folder` Text. Default: unset. The migrations folder, relative to the repository root; without it, the checks use the migration files of this configuration. ## \[secrets] The env example files, and the functions in your code that read environment variables. ### `secrets.env_examples` List. Default: `[".env.example",".env.template",".env.sample",".dev.vars.example"]`. The env example filenames to search for environment keys; a key may appear in any one of them. ### `secrets.environment_functions` List. Default: `[]`. Functions of this project that read environment variables, such as read\_env. The env-examples check reads the keys they take. ## \[site] How the site builds, where the built files go, and the size limits of its files. ### `site.build_command` List. Default: `["npm","run","build"]`. The command that builds the site. Without it, gspot runs the build script of the package manager. ### `site.build_folder` Text. Default: `"dist"`. The folder the build writes. ### `site.allowed_unlisted_pages` List or a reasoned value `{ value = …, reason = "…" }`. Default: `["404.html"]`. Reason needed when: adding. Built pages the sitemap leaves out on purpose. ### \[\[site.size\_limit]] Size limits: the built paths, and the most kilobytes they may weigh compressed. #### `site.size_limit.paths` List. Required. Built files, such as `docs/dist/_astro/*.js`; the size check runs only when a record exists. #### `site.size_limit.kilobytes` Whole number. Required. The most they may weigh together, compressed. ## \[spelling] The language of the spelling check, such as en-us or en-gb. ### `spelling.locale` `"en"`, `"en-us"`, `"en-gb"`, `"en-au"`, `"en-ca"`. Default: `"en-us"`. The English locale typos checks against: en, en-us, en-gb, en-ca, or en-au. ## \[supabase] The folders of Supabase edge functions, the schemas, and the generated types file. ### `supabase.functions_folder` Text. Default: `"supabase/functions"`. The folder, relative to the repository root, with one folder for each edge function; the default starts from the project scope. ### `supabase.schemas` List. Default: `["public"]`. The database schemas to include in generated types. ### `supabase.types_file` Text. Default: unset. The committed database types file, relative to the repository root; without it, no type comparison runs. ## \[translations] The base locale and the folder of message files. ### `translations.base_locale` Text. Default: `"en"`. The locale whose message keys every other locale must contain. ### `translations.messages_folder` Text. Default: unset. The folder of the `.json` message files, relative to the repository root. ## \[swift] The Xcode project, scheme, and destination that the Swift build uses. ### `swift.xcode_project` Text. Default: unset. The Xcode project or workspace, relative to the repository root, which init fills in when it finds one; without it, the checks use the Swift package. ### `swift.xcode_scheme` Text. Default: unset. The shared scheme the build, the analyzer, and Periphery use. ### `swift.xcode_destination` Text. Default: `"generic/platform=iOS Simulator"`. The xcodebuild destination. ## \[repository] The largest file that Git may track, in kilobytes. ### `repository.max_file_kilobytes` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `1024`. Reason needed when: raising. The largest file, in kilobytes, that Git may track without Git LFS or a linguist-generated or linguist-vendored mark in .gitattributes. ## \[prose] The longest sentence, list item, and paragraph that the prose check accepts. ### `prose.max_sentence_words` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `30`. Reason needed when: raising. The most words one sentence of documentation may have. ### `prose.max_list_item_words` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `45`. Reason needed when: raising. The most words one list item may have. ### `prose.max_paragraph_sentences` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `6`. Reason needed when: raising. The most sentences one paragraph may have. ## \[duplication] The smallest copy, in lines and tokens, that the duplication check reports. ### `duplication.min_lines` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `8`. Reason needed when: raising. The fewest lines a copied block has before it counts. ### `duplication.min_tokens` Whole number or a reasoned value `{ value = …, reason = "…" }`. Default: `40`. Reason needed when: raising. The fewest tokens a copied block has before it counts. ## \[xcode] The entitlements that an Xcode project may request. ### `xcode.allowed_entitlements` List. Default: `[]`. The entitlements the app may hold; empty turns the check off. ## \[tools] Options for single tools, such as lint rule options. gspot writes them into the tool files. ### \[tools.hadolint] #### `tools.hadolint.trusted_registries` List. Default: `[]`. The registries a base image may come from; empty allows any. ### \[tools.nginx] #### `tools.nginx.image` Text. Default: `"nginx:1.29.3-alpine"`. The image nginx -t runs in; name the one the deployment uses. ### \[tools.ruff] #### `tools.ruff.docstring_convention` `"google"`, `"numpy"`, `"pep257"`. Default: unset. The docstring style that Ruff and pydoclint check: google, numpy, or pep257. ### \[tools.semgrep] #### `tools.semgrep.registry_packs` List. Default: `["p/owasp-top-ten"]`. The registry packs the manual registry check runs. #### `tools.semgrep.rule_files` List. Default: `[]`. Rule files or folders of the repository that Semgrep runs beside the shipped packs. ### \[tools.sqlfluff] #### `tools.sqlfluff.dialect` Text. Default: `"ansi"`. The SQL dialect that SQLFluff parses, such as postgres. The postgres configuration sets it. ### \[tools.swiftlint] #### `tools.swiftlint.always_keep_imports` List. Default: `[]`. Imports the analyzer keeps although it sees no use of them. ### \[tools.trivy] #### `tools.trivy.severity` List. Default: `["HIGH","CRITICAL"]`. The severities Trivy reports. ### \[tools.eslint] #### `tools.eslint.extra` A reasoned value `{ value = …, reason = "…" }`. Default: unset. Extra options passed to the tool as written. #### `tools.eslint.import_extension` `"js"`, `"ts"`, `"extensionless"`. Default: `"js"`. The import suffix in this scope: js for compiled ESM, ts for Deno and Bun, or extensionless for code that a bundler builds. #### `tools.eslint.rules` Table. Default: `{}`. Option arrays by ESLint rule ID, without a severity, which apply only where the level turns the rule on; accept findings with gspot ignore `--rule`. #### \[\[tools.eslint.restricted\_imports]] Imports that this repository refuses, each with a name and a message. ##### `tools.eslint.restricted_imports.name` Text. Required. The import to refuse. ##### `tools.eslint.restricted_imports.message` Text. Required. What to import instead. #### \[\[tools.eslint.override]] Lint rule options for repository-relative paths, in order, with ignores applied last. ##### `tools.eslint.override.paths` List. Required. The files that these lint rule options apply to. ##### `tools.eslint.override.rules` Table. Default: unset. Lint rule options for these files. ##### `tools.eslint.override.runtime` One of 59 names. Default: unset. The runtime of these files, such as `serviceworker` (the name the `globals` package uses). ### \[tools.commitlint] #### `tools.commitlint.types` List. Default: `["feat","fix","refactor","perf","docs","test","build","ci","chore","style","revert"]`. The commit types the type-enum rule allows. #### `tools.commitlint.scopes` List. Default: `[]`. Commit scopes that the scope-enum rule accepts, and a message may leave out its scope. When empty, gspot uses the last segment of each project scope path, or no list without project scopes. #### `tools.commitlint.rules` Table. Default: `{}`. Options for active commitlint rules, by rule name: an always/never condition and an optional value, without a severity; the level decides which rules run. Ignore a finding with gspot ignore commits/commitlint `--rule` `` `--reason` ``. ### \[tools.yamllint] #### \[tools.yamllint.rules] Option tables for active YAML rules; the level decides which rules run, and \[format] sets the width. Ignore a finding with gspot ignore files/yamllint `--rule` `` `--reason` ``. ##### \[tools.yamllint.rules.indentation] ###### `tools.yamllint.rules.indentation.level` `"error"`. Default: unset. ###### `tools.yamllint.rules.indentation.ignore` Value. Default: unset. ###### `tools.yamllint.rules.indentation.ignore-from-file` Value. Default: unset. ###### `tools.yamllint.rules.indentation.spaces` `"consistent"`. Default: unset. ##### \[tools.yamllint.rules.\] ###### `tools.yamllint.rules..level` `"error"`. Default: unset. ###### `tools.yamllint.rules..ignore` Value. Default: unset. ###### `tools.yamllint.rules..ignore-from-file` Value. Default: unset. ### \[tools.markdownlint] #### `tools.markdownlint.rules` Table. Default: `{}`. Option tables by Markdown rule ID, such as MD024; the level decides which rules run. Ignore a finding with gspot ignore markdown/markdownlint `--rule` `` `--reason` ``. ### \[tools.stylelint] #### `tools.stylelint.rules` Table. Default: `{}`. Options for the Stylelint rules that run at the level, which stay errors. Accept a finding with gspot ignore css/stylelint `--rule` `` `--reason` ``. ### \[tools.squawk] #### `tools.squawk.assume_in_transaction` `true` or `false`. Default: `false`. Whether the migration runner wraps every migration in a transaction; the supabase configuration turns it on. ### \[tools.prettier] #### `tools.prettier.extra` A reasoned value `{ value = …, reason = "…" }`. Default: unset. Extra options passed to the tool as written. ### \[tools.editorconfig-checker] #### `tools.editorconfig-checker.extra` A reasoned value `{ value = …, reason = "…" }`. Default: unset. Extra options passed to the tool as written. These options go into .editorconfig. ### \[tools.typos] #### `tools.typos.extra` A reasoned value `{ value = …, reason = "…" }`. Default: unset. Extra options passed to the tool as written. ### \[tools.shellcheck] #### `tools.shellcheck.extra` A reasoned value `{ value = …, reason = "…" }`. Default: unset. Extra options passed to the tool as written. ### \[tools.knip] #### `tools.knip.extra` A reasoned value `{ value = …, reason = "…" }`. Default: unset. Extra options passed to the tool as written. #### `tools.knip.entry` List. Default: `[]`. Entry points knip starts from beyond the ones the framework configuration declares. #### `tools.knip.ignore_dependencies` List or a reasoned value `{ value = …, reason = "…" }`. Default: `[]`. Reason needed when: adding. Exact package names that knip does not report as unused; new names need a reason, and a scope list applies to that project only. ### \[tools.purgecss] #### `tools.purgecss.safelist` List or a reasoned value `{ value = …, reason = "…" }`. Default: `[]`. Reason needed when: adding. Class names a script adds at run time: the names and the reason. ### \[tools.v8r] #### \[\[tools.v8r.schema]] File patterns mapped to schema URLs, for files that the SchemaStore catalog does not know. ##### `tools.v8r.schema.paths` List. Required. Files SchemaStore does not know. ##### `tools.v8r.schema.url` Text. Required. The JSON Schema for them. ### \[tools.taplo] #### `tools.taplo.extra` A reasoned value `{ value = …, reason = "…" }`. Default: unset. Extra options passed to the tool as written. ## \[\[import\_rule]] Groups of files in import order: code may import the files of its own rule and of every rule listed above it. ### `import_rule.name` Text. Required. The name that other import rules use in can\_also\_import. ### `import_rule.paths` List. Required. The files of the import rule. ### `import_rule.can_also_import` List. Default: `[]`. The import rules listed below this one that its files may import. ## \[\[ignore]] Checks and lint rules turned off for some paths or everywhere, each with a reason. ### `ignore.check` Text. Required. The check whose findings you accept: a shipped check or a command check. ### `ignore.rule` Text. Default: unset. One lint rule of that check. ### `ignore.paths` List. Default: unset. Where the ignore applies. ### `ignore.reason` Text. Required. Why you accept the findings. ### `ignore.until` Text. Default: unset. Stop applying this ignore on this date, in YYYY-MM-DD format (UTC). ## \[\[command\_check]] Your own commands, run as checks with a stage, a file list, and an output format. ### `command_check.command` List. Required. The executable and arguments to run on selected files. ### `command_check.stage` `"commit"`, `"push"`, `"manual"`. Required. The earliest stage that runs this check. ### `command_check.summary` Text. Default: unset. A short description shown in check listings. ### `command_check.help` Text. Default: unset. Instructions for resolving findings. ### `command_check.fix` List. Default: unset. The executable and arguments that correct findings. ### `command_check.exit_codes` List. Default: unset. Nonzero statuses that mean findings from command and fix. Other nonzero statuses mean execution failure. ### `command_check.name` Text. Required. The check ID that `--only` and `[[ignore]] check` use, which may not match a shipped check. ### `command_check.paths` List. Required. Repository-relative glob patterns selecting inputs. ### \[command\_check.output] The parser and location patterns used to read findings. #### `command_check.output.format` One of these values: | Value | Meaning | | ---------------- | ------------------------------------------------------------------------------------------------------- | | `"regex"` | Read findings from lines matched by pattern and its named file, line, column, rule, and message groups. | | `"grouped"` | Read a file heading matched by file\_pattern, followed by findings matched by pattern. | | `"sarif"` | Read SARIF results, rule metadata, source locations, and fixes. | | `"knip"` | Read Knip JSON issue categories and source positions. | | `"semgrep"` | Read Semgrep JSON results, locations, rule IDs, messages, and fixes provided by Semgrep. | | `"trufflehog"` | Read TruffleHog JSON findings while withholding raw secret output. | | `"typos"` | Read typos JSON and convert byte offsets to character columns. | | `"markdownlint"` | Read markdownlint result objects and their rule and source locations. | | `"json"` | Read JSON findings using items, children, and fields to select report entries. | | `"lines"` | Report each nonempty output line as a finding without a source location. | | `"none"` | Use the exit status without parsing output into findings. | Required. How gspot reads the output: json, regex, grouped, lines, none, or the format of a named tool. #### `command_check.output.items` Text. Default: unset. For json: the path to the list of findings. #### `command_check.output.children` Text. Default: unset. For json: the field of each item that holds its findings. #### `command_check.output.line_base` `0` or `1`. Default: unset. Whether the tool counts lines from 0 or 1. #### `command_check.output.file_type` `"path"`, `"link"`, `"history"`. Default: `"path"`. What the file field holds: a path in the repository, a link target, or a path in a past commit. #### `command_check.output.pattern` Text. Default: unset. For regex: the pattern for one finding, with named groups file, line, column, rule, and message. #### `command_check.output.file_pattern` Text. Default: unset. For regex: the pattern for a line that names the file of the findings below it. #### `command_check.output.fixable` Text. Default: unset. For regex: the pattern that marks a fixable finding. #### `command_check.output.message` Text. Default: unset. For regex: the pattern for the message. #### \[command\_check.output.fields] ##### `command_check.output.fields.file` Text. Default: unset. For json: the field that holds the file. ##### `command_check.output.fields.line` Text. Default: unset. For json: the field that holds the line. ##### `command_check.output.fields.column` Text. Default: unset. For json: the field that holds the column. ##### `command_check.output.fields.rule` Text. Default: unset. For json: the field that holds the lint rule. ##### `command_check.output.fields.message` Text. Default: unset. For json: the field that holds the message. ##### `command_check.output.fields.fixable` Text. Default: unset. For json: the field that is set when the tool can fix the finding. ## \[structure] Whether files may re-export the exports of other files, and which files may read environment variables. ### `structure.reexports` `"none"`, `"index-only"` or a reasoned value `{ value = …, reason = "…" }`. Default: `"none"`. Reason needed when: adding. Where re-exports may go: none, or an index file that re-exports other files. ### `structure.environment_readers` List. Default: `[]`. The files that may read environment variables; other files get the values passed in. ## \[hooks] Whether gspot install turns on the Git hooks. ### `hooks.enabled` `true` or `false`. Default: `false`. Install and run the configured Git hooks. ## \[ci] Whether gspot writes a CI workflow, for which provider, and on which systems it runs. ### `ci.enabled` `true` or `false`. Default: `false`. Whether gspot writes a CI workflow; set it at the root only. ### `ci.provider` `"github"`, `"gitlab"`. Default: `"github"`. Which CI service gets the workflow; set it at the root only. ### `ci.setup_command` List. Default: unset. A command that CI runs before `gspot install`, such as `npm ci`; set it at the root only. ### `ci.operating_systems` List. Default: `["linux"]`. The operating systems of the CI jobs. GitHub only; refused with provider = “gitlab”. ## \[rules] Whether gspot writes rules for coding agents, and which rule files it leaves out. ### `rules.enabled` `true` or `false`. Default: `true`. Write the rules and the AGENTS.md block. ### `rules.exclude` List. Default: `[]`. Rule files or folders under .gspot/configurations/ to leave out, such as .gspot/configurations/general/engineering/rules/PLANNING.md.