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/kleosr/kleosrulesWrote 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/rules/kleosr/kleosrules/core)<a href="https://agentmods.dev/rules/kleosr/kleosrules/core"><img src="https://agentmods.dev/badge/rules/kleosr/kleosrules/core/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/rules/kleosr/kleosrules/core"><img src="https://agentmods.dev/badge/rules/kleosr/kleosrules/core.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 | $0.00866 | $0.00866 |
| Opus 5 | $0.00433 | $0.00433 |
| Sonnet 5 | $0.00173 | $0.00173 |
| Haiku 4.5 | $0.00087 | $0.00087 |
Grade A, and why
core 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 today.
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 — 44 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Core
Understand the files you will change. Change them. Verify the behavior this change can break. Correct failures. Do not claim done on stale evidence. Cite command + exit (or a driven UI path). A missing or nonzero observation is not success.
Change
State the intended behavior and non-goals when the task leaves them open. Trace callers and contracts before changing a boundary. Preserve unrelated work. If the same check fails twice, stop repeating it: record the evidence, change the hypothesis, or name the missing input. Remove temporary instrumentation when the investigation ends.
Craft
Readable before clever. Explicit before magical. Simple before abstract. Meaningful names. Predictable control flow. Comments explain decisions, not the obvious. Handle errors on purpose — no empty catch, no swallowed exit. Clarity, correctness, and changeability outrank line count. Do not write decorative abstractions, extra wrappers, or “AI-looking” patterns to satisfy a style. Domain-hard complexity is allowed when justified.
Architecture
Own one job per module. Stable public interfaces; change internals before widening the contract. Minimal coupling and duplication. Match 1–2 siblings before Write. A local fix must not silently change ownership, layering, or call direction elsewhere. No speculative infrastructure, dead compatibility layers, or drive-by files. Named exports by default (framework-required default exports win). Remove dead code this task introduced or obsoleted. Cleanup is limited to the task.
Ladder
Lowest sufficient rung when equally suitable. Choose for correctness and fit, not golf.
- No implementation change (config, delete, existing API)
- Existing repository code (Grep)
- Stdlib / platform APIs
- Installed dependencies (new dependency needs a one-line why)
- Small clear local implementation
Size
Soft ~80 lines is a cohesion preference, not golf. Split when a module holds two jobs — not to hit a count. New hand-written files: hard 300; never 500. Legacy >700: localized fixes; extracted modules ≤300. Generated, vendor, lockfiles, migrations, and declarative config are excluded.
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.
- today Changed · +3 lines · +23 tokens per session 3a744b7c7cb6
- yesterday First seen · 41 lines · 843 tokens per session scan A d9bb291a7830
core is a cursor rule published in the GitHub repository kleosr/kleosrules (0 stars, last pushed today), licensed MIT. It adds 866 tokens to every session, about $0.0043 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 2026-09-13.
Other cursor rules, from other repositories
nextjs
Rules for Next.js projects (App Router and Pages Router).
php
Rules applied when working in PHP files (incl. Laravel, Symfony, WordPress).
vue
Rules for Vue (2 and 3) single-file components.
shell
Rules for shell scripts (bash, zsh, sh, PowerShell).
tests
Rules for test files — keep tests lean and pattern-matched.
20-routing
Routing manifest rules — single source app/routes.ts. See rules/core/routing.md.