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 instructions/pinebasedev/svelteflare/agents-mdgit clone --depth 1 https://github.com/pinebasedev/svelteflareWhat 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 | $0.00874 | $0.00874 |
| Opus 5 | $0.00437 | $0.00437 |
| Sonnet 5 | $0.00175 | $0.00175 |
| Haiku 4.5 | $0.00087 | $0.00087 |
Grade A, and why
svelteflare AGENTS.md 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 3d 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 — 38 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent Rules for This Project
This is a Svelteflare app: a pnpm/Turbo monorepo. apps/web is a SvelteKit SPA that contains all UI and routing. apps/api is a Hono Cloudflare Worker that contains all server logic (auth, database, billing). apps/marketing is a small prerendered SvelteKit site (the public marketing page) that follows the same theme rules. packages/ui (imported as @repo/ui) is the theme and component library — the single source of truth for how anything looks.
Theme fidelity (non-negotiable)
The app has a founder-approved theme. All UI you build must look like it belongs to that theme.
- Build ALL UI from
@repo/uicomponents styled with semantic theme tokens only:bg-background,text-foreground,bg-primary,text-primary-foreground,bg-card,text-muted-foreground,border-border,bg-destructive, opacity modifiers liketext-foreground/60, and the other tokens defined inpackages/ui/src/global.css. - NEVER use:
- Tailwind palette classes:
bg-blue-500,text-zinc-900,text-green-500,bg-white,text-black, … - Color literals anywhere: hex (
#1e40af),rgb(),hsl(),oklch()— in classes,style=attributes, or<style>blocks - Arbitrary color values:
bg-[#...],text-[oklch(...)],border-[rgb(...)] - Named CSS colors (
color: red) or classes for tokens that don't exist inglobal.css
- Tailwind palette classes:
- Dark mode works automatically through the tokens. Do not write
dark:color overrides in feature code. - Never recreate buttons, inputs, selects, textareas, dialogs, or other controls with raw markup. Never import
bits-ui,formsnap, or icon packages directly — icons come from~icons/lucide/<name>via unplugin-icons. - Before assuming a component is missing, read the export list in
packages/ui/src/index.ts(~57 components exist). Before assuming a token is missing, read the:rootblock ofpackages/ui/src/global.css. If a primitive truly is missing:pnpm --filter @repo/ui ui:add <name>, export it frompackages/ui/src/index.ts, import it via@repo/ui. packages/ui/src/global.cssandpackages/ui/src/componentsare THEME-OWNED. Do not edit them during feature work. Change them only when the user explicitly asks to change the theme or brand, through the theme workflow — and then use the exact colors the user provides, never colors you invented.- Lint enforces the color rules (
theme/no-hardcoded-colors). If lint flags a color, replace it with a token — do not suppress the error. Run checks (just check) before finishing.
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.
- 3d ago First seen · 38 lines · 874 tokens per session scan A 63b1ecb5de57
svelteflare AGENTS.md is an instructions file published in the GitHub repository pinebasedev/svelteflare (47 stars, last pushed 1mo ago), licensed MIT. It adds 874 tokens to every session, about $0.0044 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 instructions, from other repositories
flarestarter AGENTS.md
Instructions for FlareStarter/flarestarter, covering agents.md, what this is, structure, conventions and commands.
flarestarter CLAUDE.md
Instructions for FlareStarter/flarestarter: This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
rewind AGENTS.md
Instructions for pdugan20/rewind, covering rewind, work modes, development commands, agent configuration and architecture overview.
nextjs-expo-saas-starter CLAUDE.md
Instructions for brainbytes-dev/nextjs-expo-saas-starter, covering claude.md, project overview, monorepo structure, commands and database (from root or packages/db).
cf-multiworker-starter-kit AGENTS.md
AGENTS.md instructions for firtoz/cf-multiworker-starter-kit, covering agent instructions, read first (by task), cloud agents, one-liner defaults and creating new skills.
codeguide-starter-fullstack CLAUDE.md
Claude Code instructions for CodeGuide-dev/codeguide-starter-fullstack, covering claude code task management guide, documentation available, mandatory task management workflow, step 1: discover tasks (mandatory) and step 2: start each task (mandatory).