humanise

humanise is a skill for Claude Code, Codex from Nisus74/humanise. It costs 112 tokens per session (2,392 once invoked), scanned A, original, MIT.

A writing and editing skill for producing human-facing text in a specific person's voice.

In plain words
What is it for?
It is for drafting, rewriting, reviewing, or checking emails, reports, documentation, proposals, posts, applications, and other text meant to be shared.
Why use it?
It helps preserve the writer's meaning and recognizable style while reducing generic or artificial wording.

Skill for Claude CodeCodex

Part of the humanise plugin — 1 skill, 7 commands, 5 agents, 2 hooks shipped together

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/nisus74/humanise/skill
Any agent
npx skills add Nisus74/humanise --skill skill
Clone the repo
git clone --depth 1 https://github.com/Nisus74/humanise

Made for: Claude Code, Codex.

Or install humanise, the plugin that ships this one along with the rest of its 1 skill, 7 commands, 5 agents, 2 hooks.

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 humanise

README.md
[![agentmods](https://agentmods.dev/badge/skills/nisus74/humanise/skill.svg)](https://agentmods.dev/skills/nisus74/humanise/skill)
Your own site
<a href="https://agentmods.dev/skills/nisus74/humanise/skill"><img src="https://agentmods.dev/badge/skills/nisus74/humanise/skill.svg" alt="Measured on agentmods" height="20"></a>
Per session 112 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,392 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 $0.00112 $0.02392
Opus 5 $0.00056 $0.01196
Sonnet 5 $0.00022 $0.00478
Haiku 4.5 $0.00011 $0.00239

Measured 3d ago against content hash 2d322661a7de, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

humanise 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 3d ago.

The scan reads SKILL.md. This mod also ships 7 executable files (evals/assertions/mine_weaknesses.py, evals/assertions/pairwise_trial.py, evals/assertions/run_all.py, …), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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.

skill/SKILL.md · 252 lines

How it starts

The opening of the file, as written. The whole thing — 252 lines — stays where its author put it; the contents beside it link to each section on GitHub.

humanise

Produce writing a particular person would recognise as theirs. Preserve what they mean, model the reader, draft from direct evidence of the user's voice, then remove generic model habits. Clean prose is the floor. Voice fidelity is the goal.

Operating principles

Use this order when instructions compete:

  1. The user's current request.
  2. Meaning, factual fidelity and required wording.
  3. Direct samples from the same channel and relationship.
  4. Confirmed cross-channel voice evidence.
  5. The user's soul and absolute rules.
  6. Channel, register and dialect defaults.
  7. Generic anti-AI guidance.

Never change a claim to make the prose smoother. Never invent a fact, opinion, anecdote, number, name or degree of certainty to make writing seem more human. Never manufacture quirks to satisfy a style quota.

Modes

The host may invoke the skill explicitly or select it from the request. Interpret these optional mode words after the skill name:

  • init: run progressive voice setup.
  • guide: draft new material.
  • rewrite: edit existing material.
  • check: inspect a draft for mechanical and structural tells.
  • fingerprint: rebuild the voice model after new samples.
  • learn: capture the difference between a proposed draft and the user's final version.
  • improve: run the engine improvement workflow.

Read the matching file in commands/ when a mode is named. If none is named, infer guide, rewrite or check from the request.

Boundaries

Apply the skill to prose the user will send, publish, present or approve. Hold off on:

  • source code, configuration, identifiers and commands;
  • quoted material and supplied testimonials;
  • data tables and machine-readable formats;
  • fixed legal, regulatory or contractual wording;
  • the user's raw notes when they want those preserved as source material;
  • conventional commit subjects.

Edit only the prose around exempt material. Preserve quoted spans exactly. Ask only when a wrong assumption would materially change the result.

Read the full file on GitHub · 252 lines

Files

What ships with it

58 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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. 3d ago First seen · 252 lines · 112 tokens per session scan A 2d322661a7de

Subscribe to this mod's changes

humanise is a skill published in the GitHub repository Nisus74/humanise (1 stars, last pushed 7d ago), licensed MIT. It adds 112 tokens to every session and 2,392 once invoked, about $0.0006 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 2026-08-31.