ziya: Skill for Claude Code

.agents/skills/release-announcement/SKILL.md

release-announcement is a skill for Claude Code, Codex from ziya-ai/ziya. It costs 60 tokens per session (3,661 once invoked), scanned A, original, MIT.

A guide for turning a software release changelog into a short announcement for a specific audience. A changelog is a detailed record of code changes; an announcement focuses on what users will notice.

In plain words
What is it for?
It helps extract themes from release changes, rank highlights, and prepare the right message for developer or interest-focused Slack channels.
Why use it?
Technical changelogs often contain internal debugging details that are unsuitable for a public update. This separates user-facing highlights from engineering detail.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: installed under .agents/ (shared by several agents).

This is ziya-ai/ziya's own configuration. It tells Claude Code and Codex how to work on ziya itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything ziya configures →

View source ↗ ziya-ai/ziya
Reuse

Borrowing it

Nothing to install: this file belongs to ziya-ai/ziya. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/ziya-ai/ziya/main/.agents/skills/release-announcement/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/ziya-ai/ziya

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 release-announcement

README.md
[![agentmods](https://agentmods.dev/badge/skills/ziya-ai/ziya/release-announcement/github.svg)](https://agentmods.dev/skills/ziya-ai/ziya/release-announcement)
Your own site
<a href="https://agentmods.dev/skills/ziya-ai/ziya/release-announcement"><img src="https://agentmods.dev/badge/skills/ziya-ai/ziya/release-announcement/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for release-announcement

Your own site · 80×15
<a href="https://agentmods.dev/skills/ziya-ai/ziya/release-announcement"><img src="https://agentmods.dev/badge/skills/ziya-ai/ziya/release-announcement.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 60 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,661 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
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 $0.00060 $0.03661
Opus 5 $0.00030 $0.01831
Sonnet 5 $0.00012 $0.00732
Haiku 4.5 $0.00006 $0.00366

Measured 8d ago against content hash 565343608053, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

release-announcement 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 8d 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.

.agents/skills/release-announcement/SKILL.md · 325 lines

How it starts

The opening of the file, as written. The whole thing — 325 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Writing a Release Announcement

When to use

You are announcing a tagged release and need to produce a short highlight summary from a changelog. Two jobs use this skill, and they own different channels:

  • sweep (cuts the release) announces to #ziya-dev — highlights in the channel, changelog + commit link in the thread.
  • release (publishes the artifacts) announces to #ziya-interest — one prose message, no thread.

Whichever job you are in, post only to that job's channel. Availability claims differ between them; see "Availability claims" below.

Also use this when asked for "highlights", "what's new", or "what changed" for a version.

Why this needs a skill at all

A changelog and an announcement are different genres, not different lengths. Ziya's changelog entries are deliberately enormous forensic root-cause narratives written for a future engineer debugging a regression — a single entry runs 400 words about why Number.isFinite(newScale) passed on a zero-size container. That is correct for a changelog.

If you "summarize the changelog" you produce a shorter changelog. That is the failure mode. You must instead re-derive what changed from the reader's point of view, using the changelog as evidence rather than as a draft.

Run four passes, in order. Do not start writing bullets until pass 3.


Pass 1 — AGGREGATE (do this first, always)

Group entries into themes. A theme is a user-observable capability, never a file, module, or subsystem.

Aggregate only where several changelog entries describe facets of one change — the same capability written up separately because the bugs were in different files. Two entries that a user would experience as two different things are two items, and they stay two items.

  • Do NOT aggregate to shorten the message. There is no item cap (see "Hard caps"), so merging distinct work buys nothing and costs the reader a change they can no longer see. Aggregation exists to remove duplication, not volume. A release with 200 distinct user-observable changes has 200 lines.
  • N entries that are genuinely one theme produce one title naming the theme. A bare (N fixes) count is the ONLY parenthetical permitted, and only when the count is load-bearing; never enumerate the N, and never append two or three of them as examples.
  • Read git diff --stat <prev-tag>..HEAD for new top-level modules. A cluster of new files is the mechanical signal that a platform landed, and the platform is the headline — not whichever feature built on it happens to appear first in the changelog.
  • Ask of every candidate bullet: is this the trunk or a leaf? Ship the trunk and stop there — do not follow it with the leaf as an example. Naming leaves after the trunk claim is how a one-clause bullet becomes a 251-character one; if a leaf is genuinely more important than the trunk, it replaces the trunk claim rather than trailing it.

Read the full file on GitHub · 325 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. 8d ago Changed · +109 lines · +7 tokens per session 565343608053
  2. 12d ago First seen · 216 lines · 53 tokens per session scan A 77c3da2f4848

Subscribe to this mod's changes

release-announcement is a skill published in the GitHub repository ziya-ai/ziya (39 stars, last pushed today), licensed MIT. It adds 60 tokens to every session and 3,661 once invoked, about $0.0003 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 2026-08-30.