Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add aiocean/claude-plugins/plugin install aio-html-artifactsWrote 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.
[](https://agentmods.dev/skills/aiocean/claude-plugins/aio-html-explorer)<a href="https://agentmods.dev/skills/aiocean/claude-plugins/aio-html-explorer"><img src="https://agentmods.dev/badge/skills/aiocean/claude-plugins/aio-html-explorer/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.
<a href="https://agentmods.dev/skills/aiocean/claude-plugins/aio-html-explorer"><img src="https://agentmods.dev/badge/skills/aiocean/claude-plugins/aio-html-explorer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00060 | $0.04460 |
| Opus 5.5 | $0.00024 | $0.01784 |
| Sonnet 5.5 | $0.00012 | $0.00892 |
| Haiku 4.5 | $0.00006 | $0.00446 |
Grade A, and why
aio-html-explorer 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 1mo 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.
How it starts
The opening of the file, as written. The whole thing — 182 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Side-by-side HTML explorer
An explorer is one page that puts the alternatives on a shared baseline and ends with a recommendation a reader can act on now and audit later.
Reference routing
Seven rows load every time. Every other row states a condition you can answer no to — from the request text, the contract keys, the material in hand, or the validator's output. A row whose condition does not match is a row not read, and leaving it unread is the design. Measured on a plain architecture-options request against the files on disk, 2026-08-13: 190 KB if every row fires, 118 KB with these conditions applied, of which 87 KB is the always set. The references/data-story.md row postdates that measurement and carries its own cost: 12 KB when an option is plotted against a threshold or a budget, 17 KB when a criterion cell also holds a rate, measured 2026-08-13. Paths are relative to ${CLAUDE_PLUGIN_ROOT}.
| Open | When |
|---|---|
references/base-layer.html + references/layout-primitives.css |
always — paste the skeleton, paste the primitives into <style>, then diverge |
references/voices.css |
always — the rules every voice obeys, and the five-row chooser you pick from |
references/voices/<the voice you picked>.css |
always, exactly one file — the token block for the name in the contract's voice= |
references/artifact-grammar.md |
always — job, reading ladder, encoding, spine, proof placement, interaction tier, constraints, completion |
skills/aio-html-explorer/references/explorer-grammar.md |
always — options aligned on criteria, the unmeasured cell, the matrix reflow |
references/copy-craft.md |
always — the recommendation sentence and the criterion labels are what a reader compares on |
references/evidence-and-confidence.md |
always — an explorer that recommends is asserting something it did not run a command to establish |
references/encoding-and-charts.md |
three or more numeric values across time, load, or cost get plotted |
references/data-story.md |
an option's numbers are plotted against a threshold, a budget, a target, or a window the options are being judged on — read Annotate the point for the marks that put that criterion on the chart, about 12 KB of the file's 27 KB, and Frame the number when a criterion cell holds a rate, a percentage, or a projection, about 5 KB more |
references/diagrams-css.md |
an option is drawn as a named shape — a flow, a timeline, a phase strip, or a sequence |
references/diagrams.md |
an option whose shape is a topology with the medium still open, or an inline <svg> you are about to hand-write |
references/captions.md |
a <figure>, the criteria <table>, a chart, a diagram, or a screenshot goes in the page — read the one ### block for that evidence type plus The alt / figcaption / prose partition, about 4 KB of the file's 13 KB |
references/interaction-budget.md |
weights, filters, a scenario switcher, or any <script> past a JSON data island. The beforeprint listener pasted from references/base-layer.html does not count — it is mandatory whenever the page holds a <details>, so counting it would make this row unconditional |
references/microcopy.md |
the page ships a <button>, a form field, or an export action. A <summary> on its own takes one rule and no file: it names what is behind it — "Full stack trace (42 lines)", "Why we rejected the Redis approach" — because a disclosure label is announced in isolation, the way link text is |
references/typography-and-voice.md |
the request names a brand, a company style, or a look to match; or you change a value inside the voice block |
references/copy-delint.md |
the validator printed a finding whose id starts copy., or the request says the writing reads like AI |
references/worked-examples.md |
a drafted option block that would have come out the same for any comparison |
references/revising.md |
a file already exists at the target path |
references/live-collaboration.md |
the request names commenting, questioning, or deciding inside the page |
What ships with it
2 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.
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.
- 1mo ago First seen · 182 lines · 60 tokens per session scan A 9a6c19ff1141
aio-html-explorer is a skill published in the GitHub repository aiocean/claude-plugins (4 stars, last pushed 2d ago), licensed MIT. It adds 60 tokens to every session and 4,460 once invoked, about $0.0002 per session on Opus 5.5. 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.
Other skills, from other repositories
eval-hooks
Audit Claude Code and Codex hooks for validity, safety, and correctness across settings.json, hooks.json, config.toml, plugin, and skill or agent frontmatter hooks. Resolves each command, checks event names, handler types, matchers, exit-code and JSON decision strategy, and timeouts against each host's documented…
eval-skills
Audit a project or library of skills for metadata, trigger boundaries, workflow completion criteria, local resource closure, instruction economy, tool scope, effort, and routing evidence. Use before shipping skills, after importing a skill collection, or when a skill triggers too often or not at all.
pr-triage
4-phase PR backlog management with audit, deep code review, validated comments, and optional worktree setup. Use when triaging pull requests, catching up on pending code reviews, or managing a backlog of open PRs. Args: 'all' to review all, PR numbers to focus (e.g. '42 57'), 'en'/'fr' for language, no arg = audit…
source-command-audit-whitepapers
Audit version freshness, FR/EN parity, and metadata quality of all whitepapers and recap cards.
audit-agents-skills
Audit Claude Code agents, skills, and commands for quality and production readiness. Use when evaluating skill quality, checking production readiness scores, or comparing agents against best-practice templates.
design-patterns
Detect, suggest, and evaluate GoF design patterns in TypeScript/JavaScript codebases. Use when refactoring code, applying singleton/factory/observer/strategy patterns, reviewing pattern quality, or finding stack-native alternatives for React, Angular, NestJS, and Vue.