sloplint 0.4.0 → 0.5.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: f1bc057d93ebef8a56816f986a286ce160461f8d3004c037a2f3b35ab764105e
-  data.tar.gz: 7c2e243d1795b3d417a815ea3cc0b2e2abe7d5954c7cdbfbd9cfeeeee3120213
+  metadata.gz: 74566267ccb2ee1d4caec464d3098809e24f232f75872a7df2ac61ef7d583153
+  data.tar.gz: 4e3b6196aeb6604d212d00be14032e530a644d4250820424774244fe7298fa9e
 SHA512:
-  metadata.gz: a213224903d49426954ede602803f941cc11617084f5cf93fd9990e003872424fac5ef18a65e615b737eb58c5f7b816252f8fcba303d50f09d037661e98badbe
-  data.tar.gz: 48761bf8884bd38d8494a4010c706ddf9f332b21df434d3c730654ee09e0ef0a401048df14d7118999575a538d9058ec511ccb9782fab655ddacf4b41953ce2b
+  metadata.gz: 41e707ec52643d3f57abe04867d399fef77c8e567378394e4c1e0e2ca541d8aefaa2cb908b9d0e419fc9af9f0a1aedb55adce9ef87fc187628be4a90641a4809
+  data.tar.gz: ed1b87c4a32490d582100343d2e7a4b114067737de3761fa3fb768e669f398e69b8c4e3ce4cb33d3138702d533818f5dd53b2b778d8736f848ca71c8166e61f8

data/CHANGELOG.md

@@ -5,6 +5,38 @@ follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
 
 ## [Unreleased]
 
+## [0.5.0] - 2026-08-27
+
+The catalog grows from 49 rules to 50, and `check` becomes the default
+command.
+
+### Changed
+
+- A first argument that is not a command name is now taken as a path, so
+  `sloplint draft.md` and `sloplint -` scan instead of failing with
+  `unknown command`. A mistyped command fails as a missing file, still exit 2.
+
+### New rules
+
+- `trailing-significance-participle` (`structure`, `warning`): the participle
+  clause hung off the end of a sentence to say what a fact means, "the sign
+  carries both names, showcasing the range of travellers it draws". An event
+  cannot showcase anything, so the claim belongs to a narrator who never
+  appears in the text.
+
+  The verb list is closed and narrow: `highlighting`, `showcasing`,
+  `reinforcing`, `shaping`, `enhancing`, `cementing`, `solidifying`,
+  `embodying`, `fostering`, `facilitating`, `signalling`. Verbs humans write
+  in the same position -- `driving`, `representing`, `reflecting`, `marking`,
+  `contributing`, `illustrating`, `demonstrating`, `emphasising`, `echoing`,
+  `affirming` -- stay out, because they usually take a person as the subject
+  and a regex cannot see the subject. `underscoring` is left to
+  `underscores-highlights`.
+
+  Two guards drop gerund lists: a preceding -ing word means the match is a
+  middle list item, and a following comma, "and" or "or" means it is not the
+  last item. "signalling to" is the physical gesture and does not flag.
+
 ## [0.4.0] - 2026-08-26
 
 The catalog grows from 29 rules to 49. Every new rule was probed against

data/README.md

@@ -35,7 +35,7 @@ The recipe sloplint is built around, and the one an agent should use:
 cat draft.md | sloplint check --markdown -o json -
 ```
 
-`--markdown` blanks out code and URLs first, `-o json` emits the machine-readable form, and `-` reads stdin. Exit 0 means clean, 1 means notes found, anything higher is an error. A bare `sloplint` with piped stdin is the same as `sloplint check -`.
+`--markdown` blanks out code and URLs first, `-o json` emits the machine-readable form, and `-` reads stdin. Exit 0 means clean, 1 means notes found, anything higher is an error. `check` is the default command, so `sloplint draft.md`, `sloplint -`, and a bare `sloplint` with piped stdin all scan.
 
 The human-readable form drops `-o json`:
 
@@ -62,9 +62,10 @@ The brackets mark the match; the rest is there so you can see what you're fixing
 ## Commands
 
 ```
-sloplint [-o full|json] <command> [args]
+sloplint [-o full|json] [command] [args]
 
 check        scan paths (or stdin) for AI-slop tells and report notes [default]
+             anything that is not a command name is taken as a path: `sloplint draft.md`
 rules        list the rule catalog (add --json for the machine-readable form)
 explain ID   print one rule's message, rationale, and a bad/ok example
 version      print the sloplint version
@@ -128,11 +129,11 @@ An unknown id or category in `--select`/`--ignore` is a usage error (exit 2, nam
 
 ## The rule catalog
 
-49 rules across four categories. `sloplint rules` prints them; `sloplint rules --json` gives an agent the enumerable form.
+50 rules across four categories. `sloplint rules` prints them; `sloplint rules --json` gives an agent the enumerable form.
 
 - **rhetorical-tic** (36) the cadence patterns: `no-x-no-y`, `no-x-no-y-frag`, `thats-the-whole`, `thats-how-x`, `announced-takeaway`, `exact-exactly`, `load-bearing`, `you-already-know`, `sit-with-that`, `hold-onto-that`, `cleanly`, `clean-count`, `cleanest-x`, `clean-x`, `not-nothing`, `is-is` (doubled copula), `if-im-being-honest` (the candor preamble, from slopwash.com's "false intimacy"), and more.
 - **puffery** (5) Wikipedia's "signs of AI writing": `puffery-words` (vibrant, nestled, groundbreaking, in the heart of), `rich-tapestry`, `vital-role`, `stands-serves-as`, `underscores-highlights`.
-- **structure** (7) `not-just-x-but-y`, `not-x-but-y` (the bare corrective), `em-dash` (any em dash), `em-dash-overuse` (three or more in one paragraph), `question-isnt` (the corrective frame in interrogative dress), `less-about-more-about`, and `rule-of-three`.
+- **structure** (8) `not-just-x-but-y`, `not-x-but-y` (the bare corrective), `em-dash` (any em dash), `em-dash-overuse` (three or more in one paragraph), `question-isnt` (the corrective frame in interrogative dress), `less-about-more-about`, `trailing-significance-participle` (the "…, showcasing its importance" clause), and `rule-of-three`.
 - **hedging** (1) `vague-attribution`: "some critics argue," "it is widely regarded."
 
 Severity is `warning` for strong tells, `info` for weak or contextual ones. No rule currently ships at `error`; the tier is reserved for a pattern with essentially zero false-positive risk, and none has earned that yet.

data/docs/SPEC.md

@@ -141,8 +141,11 @@ check options:
   --ignore IDS         skip these rules
 ```
 
-Bare `sloplint` with piped stdin behaves as `sloplint check -`. This is the
-common agent path: `cat draft.md | sloplint check --markdown -o json -`.
+`check` is the default command. A first argument that is not a command name is
+taken as a path, so `sloplint draft.md` and `sloplint -` both scan; a mistyped
+command therefore fails as a missing file. Bare `sloplint` with piped stdin
+also behaves as `sloplint check -`. This is the common agent path:
+`cat draft.md | sloplint check --markdown -o json -`.
 
 Deliberately left out of v1 (add when a real need shows up, not before):
 `compact` output, `--min-severity`, `--max-notes`, color, `-q/--quiet`, and a
@@ -252,7 +255,7 @@ Categories (for `--select`/`--ignore` by group):
 
 - `rhetorical-tic` — the cadence patterns (the user's list below)
 - `puffery` — Wikipedia "words to watch" (boasts, vibrant, nestled, tapestry…)
-- `structure` — rule-of-three, "not just X but Y", "the question isn't X, it's Y", "less about X more about Y", em dash, em-dash overuse
+- `structure` — rule-of-three, "not just X but Y", "the question isn't X, it's Y", "less about X more about Y", the trailing significance participle, em dash, em-dash overuse
 - `hedging` — vague attribution ("some critics argue", "it is widely regarded")
 
 Severities: `warning` for strong tells, `info` for weak/contextual ones. No
@@ -318,6 +321,7 @@ Single flat rule per word-cluster, matched as whole words:
 - `not-x-but-y` — the bare corrective "is not X but Y" with no escalation word; `info`, because the corrective/concession distinction is syntactic and the pattern can only approximate it.
 - `question-isnt` — "the question isn't/is not (whether|if|how|what|why|who) X, it's/but Y"; `info`. The resolving clause is required, so a plain rhetorical question never matches; "the real question is" is excluded.
 - `less-about-more-about` — "it's/this is/that's less about X (and) more about Y", also "… than about Y"; `info`. Both halves of the frame are required, and the subject slot is limited to the pronouns.
+- `trailing-significance-participle` — comma plus a participle from a closed verb list (highlighting, showcasing, reinforcing, shaping, enhancing, cementing, solidifying, embodying, fostering, facilitating, signalling), the clause a model hangs off a sentence to say what a fact means. Guards drop gerund lists and "signalling to". `driving`, `representing`, `reflecting`, `marking`, `contributing`, `illustrating`, `demonstrating`, `emphasising`, `echoing` and `affirming` stay out: humans write them in the same position, usually with a person as the subject, and the pattern cannot see the subject. `underscoring` is left to `underscores-highlights`.
 - `rule-of-three` — three parallel comma items ending a sentence (heuristic; `info` severity, off by default via `--select` since it false-positives).
 - `em-dash` — any em dash; `info`.
 - `em-dash-overuse` — 3+ em dashes in one paragraph; `warning`.

data/lib/sloplint/cli.rb

@@ -33,9 +33,9 @@ module Sloplint
       when "version" then out.puts(VERSION); 0
       when "help"    then out.puts(parser.help); 0
       else
-        err.puts("sloplint: unknown command #{command.inspect}")
-        err.puts(parser.help)
-        2
+        # Anything else is a path or "-": `sloplint FILE`, `sloplint -`.
+        # A mistyped command lands here too, and fails as a missing file.
+        cmd_check(argv.unshift(command), opts, out:, err:, stdin:)
       end
     rescue OptionParser::ParseError => e
       err.puts("sloplint: #{e.message}")
@@ -190,10 +190,11 @@ module Sloplint
           # exit 0 = clean, 1 = notes found, >1 = error
           # each note: {path,line,column,severity,rule,category,message,excerpt,context,rationale,suggestion}
 
-          usage: sloplint [-o full|json] <command> [args]
+          usage: sloplint [-o full|json] [command] [args]
 
           commands:
             check        scan paths (or stdin) for AI-slop tells and report notes [default]
+                         a first argument that is not a command name is taken as a path
             rules        list the rule catalog (add --json for the machine-readable form)
             explain ID   print one rule's message, rationale, and a bad/ok example
             version      print the sloplint version

data/lib/sloplint/rules.rb

@@ -1358,6 +1358,84 @@ module Sloplint
                  "half is chosen for the contrast, not because a reader proposed it, and " \
                  "the second half stands on its own."
     ),
+    Rule.new(
+      id: "trailing-significance-participle",
+      category: "structure",
+      severity: "warning",
+      # The verb list is closed and short on purpose. Wikipedia's "signs of AI
+      # writing" names eight watch words for this construction; half of them
+      # did not survive probing. In 1.85M words of pre-2022 Hacker News and
+      # 2.66M words of public-domain prose, "driving" appears in this position
+      # 14 times, "representing" 16, "reflecting" 10, "marking" 5 -- all of it
+      # ordinary English ("driving the price down", "representing a majority of
+      # the voting power"), and all of it the same shape a model produces, so
+      # no narrowing separates the two. Those verbs are out. The ones that ship
+      # are used in both corpora but never here: "highlighting" appears 17
+      # times in the HN sample and not once after a comma. "underscoring" is
+      # left out because underscores-highlights already flags it bare.
+      #
+      # "emphasizing", "echoing" and "affirming" are out for the same reason,
+      # found late: the shipped pattern flagged four sentences in the
+      # public-domain corpus and all four had a person as the subject. A person
+      # can emphasize or echo something; an event cannot. That is the animacy
+      # distinction the construction really turns on, and a regex cannot see
+      # it, so the verbs that carry a human subject in ordinary prose stay out
+      # rather than being approximated. The narrowed list flags nothing in
+      # either corpus.
+      #
+      # "ensuring" is also out, despite scoring 1 hit in each corpus: neither
+      # corpus contains software documentation, where "the mutex is held,
+      # ensuring no two writers collide" states a real consequence rather than
+      # claiming significance.
+      #
+      # Two guards keep gerund lists out. A preceding -ing word means the match
+      # is the second item of a list ("cutting, shaping stone"), and a
+      # following comma, "and", or "or" means it is not the last ("cutting,
+      # shaping and sanding"). "signaling to" is the physical gesture.
+      pattern: /\w+(?<!ing),[ \t]+
+                (?:further[ \t]+|thereby[ \t]+|thus[ \t]+|ultimately[ \t]+)?
+                (?:highlighting|showcasing|reinforcing|shaping|enhancing|
+                   cementing|solidifying|embodying|fostering|facilitating|
+                   signall?ing(?![ \t]+to\b))
+                (?![ \t]*(?:,|and\b|or\b))
+                [ \t]+\S/ix,
+      message: "Trailing -ing clause claiming significance -- a stock AI move.",
+      suggestion: "Cut the clause, or state the consequence as its own sentence with a subject.",
+      examples_bad: [
+        "The bridge reopened in March, cementing its place in the city's skyline.",
+        "Attendance doubled that year, highlighting the appeal of the new format.",
+        "Its sign carries both languages, showcasing the range of visitors it draws.",
+        "Trade routes crossed here for centuries, shaping the food people cook.",
+        "The grant was renewed, further solidifying the lab's standing.",
+        "The team shipped every week, reinforcing the sense that momentum mattered.",
+        "The colour was mixed by hand, embodying the care the shop is known for.",
+        "The lab was refitted last year, enhancing what the group can measure.",
+        "The station sits beside the port, facilitating the movement of goods inland."
+      ],
+      examples_ok: [
+        # The verb in its ordinary position, with a person or document as subject.
+        "The report highlights the two findings that changed our minds.",
+        "She emphasises the method rather than the results.",
+        # Gerund lists: the match is a middle item, not a trailing clause.
+        "The work involves cutting, shaping and polishing the stone.",
+        "They spent the morning cutting, shaping stone for the wall.",
+        "The class covers drawing, shaping, then firing the clay.",
+        # The physical gesture, not a claim about meaning.
+        "He turned toward the bar, signaling to the waiter.",
+        # War and Peace (Maude translation, public domain): a person as the
+        # subject, which is the case the verb list cannot distinguish, so
+        # "emphasizing", "echoing" and "affirming" are excluded outright.
+        "recounted Bitski, emphasizing certain words and opening his eyes",
+        "\u201cOu-rou-rou!\u201d yelled the crowd, echoing the crash of the roof",
+        # A paragraph break is not a comma.
+        "The kiln runs hot\n\nShaping the clay comes first."
+      ],
+      rationale: "A participle hung off the end of a sentence, with a fact or an event as " \
+                 "its subject, asserts what something means without anyone having to say it. " \
+                 "An event cannot highlight or showcase anything; the claim belongs to a " \
+                 "narrator who never appears. Careful writers put the verb in its own clause " \
+                 "with a subject, or leave the significance to the reader."
+    ),
     # ── hedging ───────────────────────────────────────────────────────────
     Rule.new(
       id: "vague-attribution",

data/lib/sloplint/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Sloplint
-  VERSION = "0.4.0"
+  VERSION = "0.5.0"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: sloplint
 version: !ruby/object:Gem::Version
-  version: 0.4.0
+  version: 0.5.0
 platform: ruby
 authors:
 - Benjamin Jackson