performance

performance is a cursor rule for coding agents from mhmdreza-rafiei/agent-tools. It costs 17 tokens per session (355 once invoked), scanned A, original, MIT.

A set of rules for work where speed or resource use is an explicit requirement, such as page loading time, API response time, or data-processing throughput.

In plain words
What is it for?
Use it on user-facing features, data pipelines, or frequently executed code. It helps investigate slow queries, repeated parsing, unnecessary file reads, and other bottlenecks.
Why use it?
It prevents optimization based on guesses or without a measurable target. The rules require setting a budget, measuring the current bottleneck, changing it, and measuring again.

Cursor rule

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 rules/mhmdreza-rafiei/agent-tools/performance
Clone the repo
git clone --depth 1 https://github.com/mhmdreza-rafiei/agent-tools

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 performance

README.md
[![agentmods](https://agentmods.dev/badge/rules/mhmdreza-rafiei/agent-tools/performance.svg)](https://agentmods.dev/rules/mhmdreza-rafiei/agent-tools/performance)
Your own site
<a href="https://agentmods.dev/rules/mhmdreza-rafiei/agent-tools/performance"><img src="https://agentmods.dev/badge/rules/mhmdreza-rafiei/agent-tools/performance.svg" alt="Measured on agentmods" height="20"></a>
Per session 17 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 355 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 $0.00017 $0.00355
Opus 5 $0.00009 $0.00178
Sonnet 5 $0.00003 $0.00071
Haiku 4.5 $0.00002 $0.00036

Measured 4d ago against content hash b0ade8c5dad9, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

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 4d 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.

rules/code/performance.mdc · 42 lines

What it actually says

Performance

Applies when performance is a stated requirement or the change touches a hot path. Load for user-facing features, data pipelines, hot loops. Not every chat — premature optimization is a real cost.

Budget before optimization

  • A performance change needs a budget: LCP / TBT / CLS for web, p95 / p99 latency for APIs, throughput for pipelines. No budget → no target → no way to know when to stop.
  • If the user gave no budget, ask once. If they don't know, propose a sane default and label it as a suggestion.

Measure, then fix

  • Measure before optimizing. A guess about where time goes is usually wrong.
  • Fix the measured bottleneck, then re-measure. One bottleneck at a time.
  • The cheapest fix is usually algorithmic (a better data structure, fewer passes, an index), not a micro-optimization.

Hot paths

  • N+1 queries, repeated parsing, re-reading the same file, synchronous I/O in a loop — these are the usual suspects. Check them first.
  • Cache only what is expensive and stable. A cache invalidation bug is worse than no cache.

Do not

  • Optimize without a budget and a measurement.
  • Trade correctness for speed without flagging it (a ponytail: comment naming the ceiling is fine).
  • Add a dependency for a perf gain the stdlib already covers.
  • Micro-optimize a path that isn't hot.
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. 4d ago First seen · 42 lines · 17 tokens per session scan A b0ade8c5dad9

Subscribe to this mod's changes

performance is a cursor rule published in the GitHub repository mhmdreza-rafiei/agent-tools (5 stars, last pushed 16d ago), licensed MIT. It adds 17 tokens to every session and 355 once invoked, about $0.0001 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 2026-08-31.