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.
npx agentmods add skills/tomzx/agents/create-articlenpx skills add tomzx/agents --skill create-articlegit clone --depth 1 https://github.com/tomzx/agentsWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00084 | $0.02115 |
| Opus 5 | $0.00042 | $0.01058 |
| Sonnet 5 | $0.00017 | $0.00423 |
| Haiku 4.5 | $0.00008 | $0.00212 |
Grade A, and why
create-article scanned grade A with 1 finding 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
5. Verify all links: curl every URL in the article and replace any that return 4xx, 5xx, or connection errors. How it starts
The opening of the file, as written. The whole thing — 179 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Write Article
Produces a high-quality article from a target audience definition and a set of relevant sources. Adapts tone, depth, and structure to fit the audience.
Prerequisites
- Target audience: Who will read this article (e.g., "senior backend engineers", "product managers unfamiliar with ML", "general tech enthusiasts")
- Relevant sources: URLs, file paths, or inline content to draw from (research papers, blog posts, docs, notes, etc.)
- Topic (optional): If not inferable from sources, state the article topic explicitly
If sources are not yet gathered, run /research-article first to discover the state of the art and produce a research brief whose key sources feed this skill. If you decide not to run it, say so and give your reasoning before proceeding, so the user can course-correct.
Steps
- Read all provided sources thoroughly.
- Identify the core insight or argument the article should convey.
- Calibrate tone and depth to the target audience.
- Draft the article using the structure below, following the formatting rules.
- Verify all links: curl every URL in the article and replace any that return 4xx, 5xx, or connection errors.
- Discover related articles in the same repository (see "Finding Related Articles" below) and add a "See also" section.
- Revise for clarity, flow, and conciseness.
- Include
session_link: http://localhost:10000/?session=<id>in the frontmatter, using the current session ID from context. If no session ID is available, omit the field.
Formatting Rules
- One sentence per line: Each sentence occupies its own line in the markdown source. This produces cleaner git diffs since changes to one sentence don't affect adjacent lines. Paragraphs are separated by a blank line, with sentences within a paragraph on consecutive lines (not separated by blank lines).
- Link all sources: Every source referenced in the article body must include a verified hyperlink,
not just a textual citation (e.g.,
[Author, "Title"](https://...)not justAuthor, "Title"). - Verify links: After drafting, curl every hyperlink and replace any that return 4xx or connection errors before finalizing the article.
- Bold important statements: Highlight the most important statements, key claims, and takeaways in bold so readers scanning the article still capture the core message. Be selective: bold sentences or phrases that carry the argument, not routine sentences. Avoid bolding more than one statement per paragraph on average.
- Plain vocabulary: Prefer common, everyday words over less frequent synonyms, so the article is easy to read (for example, "use" over "utilize", "help" over "facilitate", "show" over "elucidate"). Keep technical terms where they are the precise and expected choice; the goal is to avoid uncommon or showy words, not to lose precision. Calibrate this to the audience: write simpler for general readers, and allow more domain terms for expert readers.
- Name the referent: Avoid bare referential words like "this", "that", "these", and "those" when the referent could be ambiguous, since they force the reader to backtrack and resolve what is meant. Prefer naming the thing directly. For example, write "the parser fails on nested generics" rather than "this fails on nested generics", or "that approach trades latency for throughput" rather than "that trades latency for throughput". Referential words are acceptable when the referent is unambiguous and immediately adjacent, or when repeating the noun would be clunky.
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.
- 2d ago First seen · 179 lines · 84 tokens per session scan A 79fb85dfb9a8
create-article is a skill published in the GitHub repository tomzx/agents (5 stars, last pushed 4d ago), licensed MIT. It adds 84 tokens to every session and 2,115 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
babysit-pr
Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…
imagegen
Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…