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 skills add Dragoon0x/canon --skill canon-performancegit clone --depth 1 https://github.com/Dragoon0x/canonWrote 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/dragoon0x/canon/canon-performance)<a href="https://agentmods.dev/skills/dragoon0x/canon/canon-performance"><img src="https://agentmods.dev/badge/skills/dragoon0x/canon/canon-performance/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/dragoon0x/canon/canon-performance"><img src="https://agentmods.dev/badge/skills/dragoon0x/canon/canon-performance.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 | $0.00074 | $0.02895 |
| Opus 5.5 | $0.00030 | $0.01158 |
| Sonnet 5.5 | $0.00015 | $0.00579 |
| Haiku 4.5 | $0.00007 | $0.00290 |
Grade A, and why
canon-performance 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 2mo 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 — 309 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CANON · Performance
Performance is design. A slow interface fails accessibility, fails conversion, and fails users on slow networks. The targets below are the bar.
Core Web Vitals (the only metrics that matter for ranking and UX)
| Metric | Good | Needs Improvement | Poor | Source |
|---|---|---|---|---|
| LCP (Largest Contentful Paint) | ≤ 2.5s | 2.5–4.0s | > 4.0s | web.dev |
| INP (Interaction to Next Paint) | ≤ 200ms | 200–500ms | > 500ms | web.dev |
| CLS (Cumulative Layout Shift) | ≤ 0.1 | 0.1–0.25 | > 0.25 | web.dev |
Targets are 75th-percentile across mobile devices on slow 4G. "Works on my MacBook" is irrelevant.
TTFB (Time to First Byte) — the prerequisite
| Threshold | Status |
|---|---|
| ≤ 800ms | Good |
| > 1.8s | Poor (LCP cannot be good with this TTFB) |
Bad TTFB ceilings everything else. Fix server response and CDN before anything else.
LCP — Largest Contentful Paint
Most LCP failures come from one of:
- Large unoptimized hero image
- Render-blocking CSS / JS
- Slow server (TTFB)
- Web fonts that delay text paint
Hero Image Rules
<img
src="hero-1200.webp"
srcset="hero-400.webp 400w, hero-800.webp 800w, hero-1200.webp 1200w"
sizes="100vw"
fetchpriority="high"
decoding="async"
width="1200"
height="600"
alt="..."
/>
| Attribute | Why |
|---|---|
fetchpriority="high" |
Tells browser this is the LCP candidate |
srcset + sizes |
Right size per viewport |
width + height |
Reserves space (prevents CLS) |
| Format: WebP/AVIF | 25–50% smaller than JPEG |
loading="lazy" |
NEVER for the LCP image. Lazy = late paint. |
Render-Blocking Resources
Only load critical CSS in <head>. Defer everything else.
<!-- Critical CSS inline -->
<style>/* above-the-fold styles only */</style>
<!-- Non-critical CSS deferred -->
<link rel="preload" href="/styles/full.css" as="style" onload="this.rel='stylesheet'">
<!-- Scripts deferred -->
<script src="/main.js" defer></script>
<script src="/analytics.js" async></script>
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.
- 2mo ago First seen · 309 lines · 74 tokens per session scan A 4fe69c901452
canon-performance is a skill published in the GitHub repository Dragoon0x/canon (6 stars, last pushed 6mo ago), licensed MIT. It adds 74 tokens to every session and 2,895 once invoked, about $0.0003 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 2026-08-31.
canon-performance compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| canon-performanceDragoon0x/canon | Skill | 6 | 6mo ago | MIT | A | 74 | Claude Code, Codex |
| accessibilitydiegosouzapw/awesome-omni-skills | Skill | 170 | 4mo ago | MIT | C | 119 | Claude Code, Codex |
| accessibilityericrisco/rsc-harness | Skill | 503 | 2d ago | MIT | A | 86 | Claude Code, Codex |
| Website AuditPramodDutta/qaskills | Skill | 245 | 4d ago | MIT | A | 33 | Claude Code, Codex |
| huawei-cloud-cts-trace-managementhuaweicloud/huaweicloud-devkit | Skill | 70 | yesterday | Apache-2.0 | A | 249 | Claude Code |
| accessibility-auditoyi77/1ai-skills | Skill | 17 | yesterday | MIT | A | 29 | Claude Code |
| a11y-auditgreglas75/zuvo | Skill | 7 | 3d ago | MIT | A | 147 | Claude Code |