version-numbers

version-numbers is an agent for coding agents from NickZaitsev/ru-normalizr. It costs 0 tokens per session (566 once invoked), scanned A, original, MIT.

Rules for deciding whether dotted numbers are software versions or decimal fractions in Russian text normalization.

In plain words
What is it for?
Use them when ordering or changing number-processing stages and when handling versions, model numbers, measurements, and units.
Why use it?
They prevent version numbers such as 3.12.1 from being split or mistakenly read as decimals.

Agent

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 agents/nickzaitsev/ru-normalizr/version-numbers
Clone the repo
git clone --depth 1 https://github.com/NickZaitsev/ru-normalizr

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 version-numbers

README.md
[![agentmods](https://agentmods.dev/badge/agents/nickzaitsev/ru-normalizr/version-numbers.svg)](https://agentmods.dev/agents/nickzaitsev/ru-normalizr/version-numbers)
Your own site
<a href="https://agentmods.dev/agents/nickzaitsev/ru-normalizr/version-numbers"><img src="https://agentmods.dev/badge/agents/nickzaitsev/ru-normalizr/version-numbers.svg" alt="Measured on agentmods" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 566 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 $0.00000 $0.00566
Opus 5 $0.00000 $0.00283
Sonnet 5 $0.00000 $0.00113
Haiku 4.5 $0.00000 $0.00057

Measured 4d ago against content hash f6b8f8a2ef87, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

version-numbers 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 4d 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.

docs/agents/version-numbers.md · 40 lines

How it starts

The opening of the file, as written. The whole thing — 40 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Dotted numbers: version vs. fraction

normalize_versions (numerals/versions.py) runs immediately before normalize_decimals in PipelineNormalizer.run_numerals. It has to: the decimals stage matches \d+[.,]\d+ unconditionally, so whatever it sees first it claims. Any new rule about a dot between digits belongs before the decimals stage, not inside it.

What counts as a version

Two structural signals, no list of product names:

  1. Three or more components (3.12.1, 192.168.1.1). A fraction cannot have two dots, so this is unambiguous.
  2. Two components glued by a hyphen to a name (GLM-5.2, Ту-154.2).

Signal 2 needs a guard: a named value followed by a recognized unit is a measurement, not a version (X-1.5 кг), so _is_followed_by_unit bails out via resolve_unit_info. Signal 2 still catches genuine measurements written without a unit — Mach-2.5, pH-7.4 read as "два точка пять" — an accepted trade-off, since nothing structural separates them from a model number.

Why the old behaviour was broken

DECIMAL_PATTERN took 3.12 out of Python 3.12.1 and left .1 as literal text. detokenize then treated that orphaned dot as a sentence boundary, and because NUMERAL_CONTEXT_BARRIERS does not include ., get_numeral_case scanned across it, found the fraction's own сотых (ADJF plur gent) and inflected the stranded 1 to match — producing … двенадцать сотых. Одного.

The pymorphy trap behind model numbers

Ту-154.2 used to render сто пятьдесят четырёх целых. That is not a genitive rule: pymorphy parses Ту as the pronoun тот (ADJF femn sing accs), the adjective-agreement scan in _helpers.get_numeral_case picks up that case, and the accusative of четыре is syncretic with the genitive. Reading the model number as a version sidesteps it, because version components are always rendered in the nominative. Expect the same trap in any new rule that lets a number inherit case from a preceding capitalized Latin or abbreviated token.

Read the full file on GitHub · 40 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. 4d ago First seen · 40 lines · 0 tokens per session scan A f6b8f8a2ef87

Subscribe to this mod's changes

version-numbers is an agent published in the GitHub repository NickZaitsev/ru-normalizr (22 stars, last pushed 28d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 566 tokens. 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 2026-08-30.

Related

Other agents, from other repositories