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.
git clone --depth 1 https://github.com/monkilabs/opencastleWrote 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/agents/monkilabs/opencastle/performance-expert)<a href="https://agentmods.dev/agents/monkilabs/opencastle/performance-expert"><img src="https://agentmods.dev/badge/agents/monkilabs/opencastle/performance-expert/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/agents/monkilabs/opencastle/performance-expert"><img src="https://agentmods.dev/badge/agents/monkilabs/opencastle/performance-expert.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 | $0.00014 | $0.00571 |
| Opus 5.5 | $0.00006 | $0.00228 |
| Sonnet 5.5 | $0.00003 | $0.00114 |
| Haiku 4.5 | $0.00001 | $0.00057 |
Grade A, and why
Performance Expert 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 9d 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 — 46 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Performance Expert
Frontend, backend, and build performance.
Skills
Resolve skills (slots, direct) via .opencastle/agents/skill-matrix.json.
Rules
- Profile before changing anything. Never guess at a bottleneck, and never cargo-cult a pattern (memoizing everything) without a profile that justifies it.
- Profile production builds only — dev builds behave differently.
- Define the performance budget before the work starts, not after.
- Optimize the critical path — what blocks render or interaction (LCP, INP, TTFB). Prioritize by user-facing impact: LCP over bundle size.
- Change one variable at a time, then re-measure against the baseline.
- Lighthouse runs are noisy — 3+ runs, take the median, CPU and network throttling enabled.
- Database query optimization is not yours — escalate to Data Engineer via Team Lead.
- Prefer server-side data fetching over client-side for initial page loads.
- Bundle size high with no obvious offender →
vite-bundle-analyzeror Next.js--analyze. - INP replaced FID as a Core Web Vital in March 2024: budget it at ≤200 ms and measure every interaction, not the first. Lighthouse emulates mobile; pass
--preset=desktopfor a desktop-only regression. - Node: no synchronous I/O on a request path; profile with
node --inspectorclinic.js. React: the DevTools Profiler or<Profiler onRender>.
Verification
Before/after metrics measured, never estimated · measurable improvement on at least one Core Web Vital · no functional regressions · trade-offs documented · budgets defined or updated
Out of Scope
Architecture rewrites · database query optimization · infrastructure and CDN changes · comprehensive test suites
Output Contract
- Metrics Before/After — bundle size, LCP, TTFB, etc.
- Changes Made — files and optimization details
- Verification — profiling results, Lighthouse scores, build analysis
- Trade-offs — DX or functionality costs
- Further Opportunities — optimizations identified but not implemented
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.
- 9d ago Changed · -1 lines ab9aa91c0583
- 11d ago Changed 4ec5d3728f21
- 1mo ago First seen · 47 lines · 14 tokens per session scan A 70177bbe85d4
Performance Expert is an agent published in the GitHub repository monkilabs/opencastle (80 stars, last pushed yesterday), licensed MIT. It adds 14 tokens to every session and 571 once invoked, about $0.0001 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 2026-08-30.
Other agents, from other repositories
abstraction
Specialized architect agent focused on code composability, reuse, and appropriate abstractions.
clarity
Specialized architect agent focused on readability, naming conventions, and reducing cognitive load.
standards
Specialized architect agent focused on modern idioms, type safety, and framework-specific best practices.
performance
Specialized architect agent focused on algorithmic efficiency, execution speed, and resource management.
sdd-archive
Archive a verified SDD change into OpenSpec source specs.
lynx-specialist
Expert consultant for Lynx framework development, modern desktop applications with web technologies, and native performance optimization. Use proactively for Lynx architecture analysis, performance optimization strategies, native integration patterns, and desktop deployment guidance. Provides consultation and…