humanize

humanize is a skill for Claude Code, Codex from AcKeskin/contexture. It costs 73 tokens per session (1,498 once invoked), scanned A, original, MIT.

An advisory checker and editor for making user-facing writing sound more human while preserving its meaning.

In plain words
What is it for?
It helps review and revise documents, emails, pull requests, proposals, and issues for noticeable generated-text texture.
Why use it?
It identifies clusters of AI-like writing patterns without treating detection as certain or rewriting technical model instructions.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: reads .claude/ paths.

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/ackeskin/contexture/humanize
Any agent
npx skills add AcKeskin/contexture --skill humanize
Clone the repo
git clone --depth 1 https://github.com/AcKeskin/contexture

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for humanize

README.md
[![agentmods](https://agentmods.dev/badge/skills/ackeskin/contexture/humanize.svg)](https://agentmods.dev/skills/ackeskin/contexture/humanize)
Your own site
<a href="https://agentmods.dev/skills/ackeskin/contexture/humanize"><img src="https://agentmods.dev/badge/skills/ackeskin/contexture/humanize.svg" alt="Measured on agentmods" height="20"></a>
Per session 73 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,498 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00073 $0.01498
Opus 5 $0.00036 $0.00749
Sonnet 5 $0.00015 $0.00300
Haiku 4.5 $0.00007 $0.00150

Measured 2d ago against content hash 7444ba4c716b, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

humanize scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/humanize/SKILL.md · 83 lines

How it starts

The opening of the file, as written. The whole thing — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Humanize

Overview

Make user-facing prose read like a person wrote it, without changing what it says. Grounded in a verified research base (Wikipedia AI-Cleanup + peer-reviewed detection literature). The catalogue is the source of truth; this skill applies it.

Core principles (all load-bearing):

  • Signs, not proof. Scorer, not oracle. Every marker also appears in genuine human prose (LLMs trained on it). Report advisory density/likelihood — never a binary "this is AI". No text-only detector escapes a false-positive floor.
  • Density, not instance. One "delve", one em-dash, one rule-of-three is noise. Flag clusters and frequency, never a lone occurrence.
  • User-facing track only. This governs READMEs, public docs, email, PR/proposal/issue bodies. It must NOT touch the terse model corpus.
  • Collaborator. Findings are proposed; the rewrite is a draft. The catalogue grows via /capture, never by self-editing this file.

When to use

  • User types /humanize (on a pasted draft, a selection, or a file path).
  • User asks to "humanize / de-AI / make this human / does this sound like AI / voice-check this".

Refuse (scope-guard): if the target is the terse model corpus — memory bodies (~/.claude/projects/*/memory/), codemap.md, spec/plan artefacts, Claude-facing docs — stop and say:

That's model-facing corpus (compression-disciplined by design). Humanizing it is a regression — it's meant to be terse. /humanize is for user-facing prose only.

Don't use for: commit messages (own hygiene rule), or non-prose copy-editing.

Pipeline

0. Detect register

Classify as tech-doc, email, or project-internal. State the detection; allow override. Load the matching module from references/. If genuinely ambiguous, ask.

1. Length gate

If under ~120 words (classifiers) / ~200 (GPT-4-class), say so and score conservatively or aggregate (e.g. across a sender's recent messages / a PR author's recent descriptions). Never a confident per-message verdict below the floor — it's a plateau, not a cliff, but short text is unreliable.

Read the full file on GitHub · 83 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 83 lines · 73 tokens per session scan A 7444ba4c716b

Subscribe to this mod's changes

humanize is a skill published in the GitHub repository AcKeskin/contexture (2 stars, last pushed 1mo ago), licensed MIT. It adds 73 tokens to every session and 1,498 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.