aismell

Sniff AI-smell in your text. Bilingual ES/EN. CLI. Offline. Opinionated.

🇪🇸 Léeme en español · 🌐 Try it in your browser

text.md  •  47 sentences  •  smell: 31% (moderate)

Line-by-line findings
  🔴 L4   It's not just about productivity—it's about transformation.
        AI's signature negative parallelism [en.not_just]
        → state the positive idea directly

  🔴 L12  ...let's delve into the complexities of...
        the most flagged AI word of 2024 [en.delve]
        → use "explore", "look at", or just dig in

What it actually does

aismell reads your text and points out lines that smell like a language model wrote them. No false promise of "fooling GPTZero." Your blog post, email, PR description, or essay either sounds like you or sounds like an LLM — this tool just helps you spot which.

It runs offline, uses regex + heuristics (no API calls by default), and works on Spanish and English text.

Why another one of these

Most "AI text humanizers" are scams. They reshuffle synonyms or insert typos to bypass detectors. That game is dishonest and useless.

aismell is the opposite:

• Analysis does not rewrite for you. It tells you what reads like AI, line by line. (The web has a separate "strip marks" action that can rewrite if you explicitly use it; the CLI doesn't.)
• It is auditable. Every pattern lives in a YAML file you can read and contribute to.
• It is local by default. Editorial analysis and Unicode cleanup run in your browser without sending your text. Bibliography lookup is opt-in and sends only parsed references to public catalogues. Only an explicitly enabled rewrite sends the full text to aismell on Cloudflare Workers AI. The application adds no persistent text storage; that is not a guarantee about third-party infrastructure.
• It does not promise to fool detectors. It promises to make your text feel less like autopilot.

The honest framing is editorial, not adversarial. Use it before publishing.

Web: strip marks

aismell.me has a "strip marks" action next to the normal analysis:

• Cleans Unicode locally, using the same implementation as the service. Preserves meaningful emoji joiners and script characters by default.
• Rewrites only after consent: open the panel, enable rewrite, then select “Send and rewrite”. The limit is 3,000 characters. The original is never replaced automatically.
• A rewrite uses Cloudflare Workers AI and can alter facts, names or citations despite its instructions. Compare both versions before using it.
• Bibliography and experimental paragraph-similarity review are off by default. Similarity downloads a large local model; it does not modify the editorial index.
• Findings link to the exact source span for editing; editing invalidates the old report. Word uploads retain commented downloads.

Install

pipx (recommended, any OS):

pipx install aismell
pipx inject aismell pymupdf       # optional: enable PDF annotation

Arch (AUR):

yay -S aismell                    # once published
# or from this repo:
git clone https://github.com/brm-src/aismell.git
cd aismell/packaging && makepkg -si

From source, no install:

git clone https://github.com/brm-src/aismell.git
cd aismell
./bin/aismell --help
# requires: PyYAML (pacman -S python-yaml, apt install python3-yaml, etc.)

Requires Python 3.9+. PyMuPDF is optional, only for .pdf annotation.

Use

# analyze a file (auto-detects language)
aismell post.md

# pipe stdin
cat draft.txt | aismell

# annotate a Word .docx — yellow highlights + sidebar comments
aismell mi-tesis.docx --out mi-tesis-marcado.docx

# annotate a PDF — yellow highlights + sticky-note comments
aismell paper.pdf --out paper-marcado.pdf

# no arguments? opens a native file picker (if you have a desktop)
aismell

# only show high-confidence findings
aismell --strict post.md

# get a single number 0-100 (great for CI / pre-commit)
aismell --score-only post.md

# force language
aismell --lang es post.md

# disable color
aismell --no-color post.md

Conservative Python helper

The package also exposes a small offline helper for high-confidence cleanup. It does not call an LLM or send text anywhere:

from aismell.quickclean import clean

result = clean("It is important to note that the report is ready.")
print(result.text)
print(result.changes)

Use this helper when privacy matters more than a broad rewrite. The Omarchy plugin's clean and improve actions are separate online flows documented in ai quick clean.

What it catches

Three layers of detection.

1. Phrase patterns. Things AI writes when it coasts: vale la pena destacar, delve into, stands as a testament, hope this helps. The rules now cover 130+ English patterns and 180+ Spanish patterns, including formal connectors and bloated abstract nouns.

2. Structure patterns. Tells beyond word choice: split binary reframes, false agency, evasive passive voice, negative listing, copula avoidance, generic conclusions, sycophantic openers.

3. Rhythm, shape, and discourse. Sentence-length variance, em-dash density, list ratio, rule-of-three frequency, paragraph-length symmetry (LLMs produce paragraphs of suspiciously similar length), opening-connector density (Pangram research shows multiple paragraphs starting with explicit connectors is the single strongest AI tell), and StoryScope-inspired cues such as over-explained themes and too-neat narrative closure.

Scores are a heuristic editorial index from 0 to 100, not a percentage of generated text or a probability of authorship. Higher values prioritize review; the web uses the same index as the CLI without a second classifier or invented confidence. Mentioned dates do not establish document provenance. No independent accuracy benchmark is claimed.

What it does not do

• Does not call an LLM (a future --rewrite flag may do so, opt-in). The web's "strip marks" action does use a model when you enable rewrite.
• Does not rewrite your text for you in the CLI. The web's "strip marks" action can rewrite, only when you explicitly use it.
• Does not claim to be a forensic AI detector. False positives are fine — humans use these phrases too. The point is making them visible so you can decide.

Patterns are data

Every detection rule lives in patterns/es.yaml or patterns/en.yaml. Format:

- id: en.delve
  kind: phrase            # phrase | regex
  severity: 3             # editorial priority: 1 (low) | 2 (medium) | 3 (high)
  pattern: delve into
  message: the most flagged AI word of 2024
  suggestion: use "explore", "look at", or just dig in

Add a pattern, open a PR. No code changes needed for new rules. Run bash docs/sync.sh to update every web/Worker copy. Rules are cached per process; restart after editing them, or call load_patterns.cache_clear() in a long-lived Python embedding.

Roadmap

• More patterns (always)
• --rewrite flag with optional LLM backend (offline by default stays default)
• Plain-text output mode for grep-style piping
• Pre-commit hook config
• --diff mode that shows before/after suggestions

Related

This project is built on top of the patterns documented in Wikipedia: Signs of AI writing, blader/humanizer, and hardikpandya/stop-slop. The bloat lexicon and structural checks are informed by Pangram Labs research (2024-2025), the DAMAGE paper on AI humanizers, and EditLens. See docs/research.md for the full breakdown.

All sources MIT-compatible, and so is this.

License

MIT. See LICENSE.