AI Slop Detector

Before you start: read this whole install prompt, tell me in a few lines what this skill does and what installing it involves, and wait for my go-ahead before you do anything else. Prerequisites: none for the install itself: no account, no API key, no paid service, and nothing else to install first. The skill is a single Markdown file (SKILL.md) with no scripts, no dependencies and no runtime. For using the skill afterwards, read this first: it makes no network call, calls no external model or API and sends nothing anywhere by itself — the draft text I give it is processed only by you, my own agent, in our conversation, under your own account and terms (see the disclosure in point 2). A separate web tool on feedhive.com is mentioned in the repository README outside the skill folder; it is not part of this skill, this install never opens it, and I would only ever use it by pasting a draft there myself, outside this install. Install "ai-slop-detector" for me. It helps you write without AI slop, or check a draft for it: a list of 120 patterns in 9 categories (Vocabulary 20, Stock phrases 15, Sentence patterns 22, Structure and rhythm 12, Imagery and metaphor 6, Substance 12, Formatting and punctuation 10, Chatbot residue 8, Social media and LinkedIn 15). When you write or edit for me (posts, emails, articles, scripts or docs), you avoid every pattern and prefer concrete facts, names, numbers and plain verbs, and you never invent facts, examples or numbers to replace a pattern. When you check a draft, you list each pattern you find with its ID and name, the exact text it appears in, and a fix in a few words; you do not rewrite the draft unless I ask, you do not score it, and you do not guess whether AI wrote it. Repository: https://github.com/FeedHive/ai-slop-detector (the skill is the whole skills/ai-slop-detector/ folder at the pinned commit: SKILL.md only, 1 file) Pinned version: install the code at commit 95d550706040191b0e75f1d89296a648bacb126e exactly. Do not substitute a newer version. If this commit is unreachable, stop and tell me. 1. Clone https://github.com/FeedHive/ai-slop-detector into a temporary folder, check out commit 95d550706040191b0e75f1d89296a648bacb126e, and copy the skills/ai-slop-detector/ folder from that checkout. Then compare every file in the folder with the fingerprint recorded for this listing on 2026-10-19: [FINGERPRINTS] - SKILL.md: SHA-256 f47a23c5dd171df2740f3c96959a773386043c066dc5b35ac9b92d9c31ade714 If a file does not match, it changed since the listing was prepared: do not install it yet, tell me what you found and let me decide. 2. Read SKILL.md in full, and tell me in plain words: what the skill does when used, which services it connects to, whether it reads or sends any personal information or keys, and anything that seems unusual. Stop on any red flag and tell me before going further. Confirm these points from the file: - The skill folder holds 1 file only: SKILL.md (22,494 bytes), with valid frontmatter (name: ai-slop-detector). It contains no code, no scripts, no imports and no URLs at all. The skill is self-contained in its own folder and needs no file from elsewhere in the repository: patterns.json and scripts/build.mjs at the repository root are only the upstream source and generator that produced SKILL.md (build.mjs uses only node:fs, node:path and node:url locally to regenerate README.md and SKILL.md); neither is copied, run or needed at install time or at use time. SKILL.md never references them. - Nothing is sent anywhere during installation itself; the only download at install time is the skill folder from GitHub at the pinned commit. No script is run during installation, no key is asked for, and there is no telemetry in the folder. - What leaves this machine at use time, disclosed here before it ever happens: nothing, by the skill itself. Using the skill is a separate step that only starts when I ask you to write, edit or check a text. The skill adds no destination, no API and no external model call: the text I give you is processed only by you, my own agent, inside our conversation, exactly as any text I give you already is. Do not paste my draft into any website, do not open the separate feedhive.com web tool mentioned in the repository README (https://www.feedhive.com/tools/ai-slop-detector) for me, and do not send my text to any other service: that web tool sits outside the installed skill folder and is not part of this install. - The honest scope from SKILL.md itself: the list does not decide whether a person or an AI wrote a text, a single isolated hit is minor, clusters are what matter, and when a draft has no specific detail you ask me for one or keep the sentence plain instead of inventing one. - Do not run anything to use the skill during this install, do not check any draft, and do not write any sample text with it: using the skill is my decision, later, when I actually ask. 3. When I say "yes, install", install it as a skill: copy the whole ai-slop-detector folder into the agent's skills directory, in a folder named "ai-slop-detector", keeping its 1 file (SKILL.md at its root). A manual clone followed by copying the skills/ai-slop-detector folder into ~/.claude/skills/ is the route described here; any other route is fine only as long as every installed file matches its fingerprint. 4. Verify with no network calls: frontmatter valid (name: ai-slop-detector), the 1 file in place with a matching fingerprint. Do not run any usage to verify. 5. Report what was installed, where, and what I still need to do myself (for example: nothing to set up — no key, no account, no dependency; using the skill is a separate step that only starts when I ask you to write, edit or check a text, and no send disclosed above happens because the skill itself sends nothing; if I ever want the separate feedhive.com web checker, that is my own manual choice outside this skill, where I would paste the draft myself). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. If anything looks off, stop and ask me.