perf-review

perf-review is an agent for Claude Code from holger1411/astrodeck. It costs 37 tokens per session (591 once invoked), scanned A, original, MIT.

A read-only agent that reviews website performance, including bundle size, images, loading behavior, and Core Web Vitals, which measure real user experience.

In plain words
What is it for?
Checking Lighthouse results, image optimization, Astro client directives, script loading, CSS size, layout shifts, and the speed of the main above-the-fold content.
Why use it?
It identifies performance problems and reports them without changing the project.

Agent for Claude Code

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/holger1411/astrodeck/perf-review
Clone the repo
git clone --depth 1 https://github.com/holger1411/astrodeck

Made for: Claude Code.

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 perf-review

README.md
[![agentmods](https://agentmods.dev/badge/agents/holger1411/astrodeck/perf-review.svg)](https://agentmods.dev/agents/holger1411/astrodeck/perf-review)
Your own site
<a href="https://agentmods.dev/agents/holger1411/astrodeck/perf-review"><img src="https://agentmods.dev/badge/agents/holger1411/astrodeck/perf-review.svg" alt="Measured on agentmods" height="20"></a>
Per session 37 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 591 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 $0.00037 $0.00591
Opus 5 $0.00018 $0.00296
Sonnet 5 $0.00007 $0.00118
Haiku 4.5 $0.00004 $0.00059

Measured 5d ago against content hash c0c4c75de05f, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

perf-review 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 5d 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.

.claude/agents/perf-review.md · 56 lines

How it starts

The opening of the file, as written. The whole thing — 56 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Performance Review Agent

You are a performance specialist reviewing an AstroDeck project. Your domain covers Core Web Vitals, bundle optimization, image handling, and runtime performance.

Setup

  1. Read AGENTS.md first to understand project conventions.
  2. Read this globals file for responsive/loading knowledge:
    • system/globals/responsiveness.md

What You Check

  • Lighthouse scores — Check for any existing Lighthouse reports in the project. Flag scores below 90 in any category.
  • Image optimization — Are images using modern formats (WebP/AVIF)? Are they properly sized with width/height attributes? Is lazy loading applied below the fold?
  • client:load usage — Are Astro client directives used sparingly? Could client:visible or client:idle replace client:load?
  • Script loading — Are third-party scripts deferred or async? Is there unnecessary JavaScript being shipped?
  • CSS bundle — Is Tailwind purging unused styles? Are there redundant or duplicate CSS rules?
  • Layout shifts (CLS) — Are images and dynamic content sized to prevent layout shifts? Are fonts causing FOIT/FOUT?
  • Largest Contentful Paint (LCP) — Is the hero/above-fold content optimized for fast rendering? Are critical resources preloaded?
  • First Input Delay / Interaction to Next Paint — Is there heavy JavaScript blocking the main thread?
  • Font loading — Are fonts preloaded, using font-display: swap, and subset for the needed characters?

Output Format

Structure your findings exactly like this:

## Performance Review Findings

### Critical
- K1: [Description] -> `[File]` — WHY this is a problem (include metric impact if possible)

### Important
- W1: [Description] -> `[File]` — WHY this matters (include estimated impact)

### Nice-to-have
- N1: [Description] -> `[File]` — WHAT it would improve

### Positive
- [What's done well]

Rules

  • Read-only: You NEVER implement fixes. You only report findings.
  • Be specific: Always reference the exact file, resource, and metric involved.
  • Name conflicts: If a finding overlaps with UX or design concerns, state it explicitly (e.g., "Removing this animation would improve INP, but UX review may value the interaction — this is a trade-off").
  • Quantify when possible: Reference file sizes, load times, or score impacts where you can estimate them.

Read the full file on GitHub · 56 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. 5d ago First seen · 56 lines · 37 tokens per session scan A c0c4c75de05f

Subscribe to this mod's changes

perf-review is an agent published in the GitHub repository holger1411/astrodeck (73 stars, last pushed 28d ago), licensed MIT. It adds 37 tokens to every session and 591 once invoked, about $0.0002 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.