GitHub - yasyf/slop-cop: SlopCop Claude Code / Cursor Plugin

A Go CLI that detects the rhetorical and structural tells of LLM-generated prose and emits a structured JSON report.

Built for agents, not humans: no TUI, no highlighting, no prompts — just JSON on stdout, diagnostics on stderr. It ships with a plug-and-play skill so a coding agent runs slop-cop on its own drafts and silently revises before replying.

Install

Most people use slop-cop through their coding agent. The skill activates whenever you ask the agent to write, edit, or polish prose.

Claude Code:

/plugin marketplace add yasyf/slop-cop
/plugin install slop-cop@slop-cop

Cursor: open the Plugins panel and install from Git URL https://github.com/yasyf/slop-cop.

On the first draft, the skill runs scripts/install-binary.sh (or .ps1 on Windows) to download the prebuilt binary for your platform from the rolling latest release. No Go toolchain required. Verify end-to-end with scripts/test-plugin.sh.

To invoke slop-cop directly from CI, pre-commit hooks, or scripts, install the binary:

brew install yasyf/tap/slop-cop

The --llm modes and rewrite additionally require the claude CLI on $PATH. slop-cop never needs an Anthropic API key — claude -p uses your Claude subscription.

Quickstart

slop-cop check article.md --pretty

{
  "text_length": 135,
  "violations": [
    {
      "ruleId": "era-opener",
      "startIndex": 0,
      "endIndex": 12,
      "matchedText": "In an era of"
    }
  ],
  "counts_by_rule": { "era-opener": 1 },
  "counts_by_category": { "rhetorical": 1 }
}

Input is the positional argument (- or omitted for stdin). The mode is picked from the file extension; pass --lang to force one (text, markdown, html, jsx, tsx, ts, js). Add --llm for Claude-backed semantic passes and --lines 50:80 to report only violations beginning in an edited range.

Commands

Command	What it does
`check [path\|-]`	Run detectors; emit a JSON violation report.
`rewrite [path\|-]`	Rewrite a paragraph via `claude -p`, optionally targeting `--rules`.
`rules`	Print the rule catalogue as JSON (`--category`, `--llm-only`).
`version`	Print build metadata as JSON.

Run slop-cop check --help for the full flag list, or slop-cop rules --pretty for the rule taxonomy. Exit codes: 0 success, 2 input/IO error, 3 claude subprocess error, 4 usage error.

How it works

check runs 35 instant client-side detectors (regex + structural, no external calls), then optionally Claude-backed passes: a sentence tier (Haiku, 10 rules) under --llm and a document tier (Sonnet, 3 rules) under --llm-deep — 48 rules total. --llm-effort (off｜low｜high｜auto) is the underlying control; --llm/--llm-deep are sugar for low/high. Under a plugin ($CLAUDE_PLUGIN_ROOT or $CURSOR_PLUGIN_ROOT set) with claude reachable, auto resolves to high; the resolved effort and per-tier outcomes are reported in the JSON, and an unreachable claude degrades gracefully (the failure lands in an error field while client-side results still return).

--lang parses code and masks every non-prose byte before detectors run, so you get hits on prose only — JSDoc, string literals, JSX text, HTML copy, markdown. Masking preserves length and newline offsets, so byte offsets in violations index the original UTF-8 input (bytes, not UTF-16 code units — account for that when slicing in JavaScript or Java). See skills/slop-cop-prose/SKILL.md for the exact agent instructions.

Development

The repo pins Go 1.26.2 via .tool-versions (run asdf install, or install Go 1.26 manually).

go test ./...
go vet ./...
go build ./...

main is the release: every push triggers release.yml, which cross-compiles and publishes an immutable v0.1.<run_number> GitHub Release marked Latest.

License & credit

The Go source in this repository is released under the MIT License.

The pattern taxonomy, rule catalogue, detector algorithms, word lists, and LLM prompts are derived from awnist/slop-cop by @awnist, which carried no open-source licence at port time; see NOTICE for provenance and compliance guidance. If you plan to use this beyond personal use, reach out to @awnist to clarify licensing of the derived content.