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/nucliweb/webperf-snippets/webperf-resourcesnpx skills add nucliweb/webperf-snippets --skill webperf-resourcesgit clone --depth 1 https://github.com/nucliweb/webperf-snippetsWhat 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 | $0.00172 | $0.01988 |
| Opus 5 | $0.00086 | $0.00994 |
| Sonnet 5 | $0.00034 | $0.00398 |
| Haiku 4.5 | $0.00017 | $0.00199 |
Grade A, and why
webperf-resources 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 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.
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 — 199 lines — stays where its author put it; the contents beside it link to each section on GitHub.
WebPerf: Resources & Network
JavaScript snippets for measuring web performance in Chrome DevTools. Execute with mcp__chrome-devtools__evaluate_script, capture output with mcp__chrome-devtools__get_console_message.
Scripts
scripts/Network-Bandwidth-Connection-Quality.js— Network Bandwidth & Connection Quality
Common Workflows
Network Quality Assessment
When the user asks about network performance, connection quality, or adaptive loading:
- Network-Bandwidth-Connection-Quality.js - Analyze network bandwidth, effective connection type, RTT, downlink, save-data mode
Adaptive Loading Strategy
When implementing adaptive loading or the user asks "how to optimize for slow connections":
- Network-Bandwidth-Connection-Quality.js - Detect connection quality
- Cross-reference with webperf-loading skill:
- TTFB.js (measure server response on slow connections)
- Find-render-blocking-resources.js (identify heavy resources to defer)
- Resource-Hints-Validation.js (optimize preconnect for slow networks)
- Cross-reference with webperf-media skill:
- Image-Element-Audit.js (implement responsive images based on connection)
- Video-Element-Audit.js (adjust video quality based on connection)
Save-Data Mode Detection
When the user asks about save-data or data-saving features:
- Network-Bandwidth-Connection-Quality.js - Check if save-data is enabled
- If save-data is enabled, recommend:
- Reducing image quality
- Disabling autoplay videos
- Deferring non-critical resources
- Using low-res thumbnails
Mobile/Slow Connection Optimization
When the user asks "optimize for mobile" or "slow connection users":
- Network-Bandwidth-Connection-Quality.js - Assess connection type
- Cross-reference with webperf-loading skill:
- TTFB.js (critical for slow connections)
- Find-render-blocking-resources.js (minimize blocking on slow networks)
- Critical-CSS-Detection.js (inline critical CSS to reduce RTT)
- Prefetch-Resource-Validation.js (avoid excessive prefetch on slow connections)
What ships with it
3 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.
- 2d ago First seen · 199 lines · 172 tokens per session scan A 88baa0db169e
webperf-resources is a skill published in the GitHub repository nucliweb/webperf-snippets (1,445 stars, last pushed 1mo ago), licensed MIT. It adds 172 tokens to every session and 1,988 once invoked, about $0.0009 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.
Other skills, from other repositories
module_bottleneck
Find slowest PyTorch modules in recent steps.
geo-technical
Technical SEO audit with GEO-specific checks — crawlability, indexability, security, performance, SSR, and AI crawler access.
evalscope
LLM evaluation & inference performance testing via the evalscope CLI. Translates natural language requests into evalscope commands for: (1) Model accuracy evaluation — runs 160+ benchmarks against local checkpoints or API endpoints (OpenAI-compatible, Anthropic, LiteLLM); (2) Performance stress testing — TTFT, TPOT…
hotpath_init
Configure hotpath profiling in a Rust project. Adds the hotpath dependency with feature-gated setup, instruments main with hotpath::main, functions with measure/measureall, and wraps channels, mutexes, rwlocks, streams, futures, reqwest clients, axum routers and byte-level I/O with hotpath macros. Use when the user…
opentelemetry-net-instrumentation
Provides guidance for implementing OpenTelemetry instrumentation in .NET codebases, covering tracing (Activities/Spans), metrics, logs, naming conventions, error handling, performance, SDK setup, resources, context propagation, and API design best practices.
hardwood-review
Review a Hardwood PR, local branch, or staged diff against the project's CLAUDE.md conventions and recurring failure modes. Use whenever the user asks to "review", "look over", "check", "audit", or "assess" a PR number, a branch, or pending changes in the hardwood-hq/hardwood repo. Also use when the user pastes a PR…