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/dyeoman2/tanstack-start-template/agents-mdgit clone --depth 1 https://github.com/dyeoman2/tanstack-start-templateWhat 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 | $0.02686 | $0.02686 |
| Opus 5 | $0.01343 | $0.01343 |
| Sonnet 5 | $0.00537 | $0.00537 |
| Haiku 4.5 | $0.00269 | $0.00269 |
Grade A, and why
tanstack-start-template 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 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 — 277 lines — stays where its author put it; the contents beside it link to each section on GitHub.
TanStack Start Agent Guide
Core Philosophy
- Strong end-to-end type safety (database → server → router → client); no
anyor type casting. - Server-first, progressively enhanced flows; UI hydrates cleanly but works without JS.
- Performance by design: static imports, parallel data fetching, targeted cache updates.
Architecture
src/routes/: File-based routes with loaders, guards, pending/error components.src/features/: Feature slices with UI, hooks, and server functions (*.server.ts).src/lib/server/: Server utilities (database, auth, email, env).src/components/ui/: Shadcn/ui primitives and shared components.- Database access via Convex queries/mutations and
setupFetchClient.
Golden Rules
- Keep imports static and synchronous inside server modules; no dynamic imports in server functions.
- One server function, one responsibility. Compose higher-level flows by orchestrating smaller server functions.
- Use Convex queries/mutations for data operations with automatic type generation.
- Reuse provided auth guards (
routeAuthGuard,routeAdminGuard,requireAuth,requireAdmin)—do not reimplement session checks. - Keep UI components pure; business logic lives in hooks or server functions.
- Never commit files with git unless explicitly requested by the user.
Key Patterns
Routes
export const Route = createFileRoute('/app')({
pendingMs: 150,
pendingMinMs: 250,
pendingComponent: ShellSkeleton,
component: AppLayout,
errorComponent: DashboardErrorBoundary,
});
function AppLayout() {
const navigate = useNavigate();
const location = useLocation();
const { isAuthenticated, isPending } = useAuth();
const redirectRef = useRef(false);
const redirectTarget = location.href ?? '/app';
useEffect(() => {
if (isPending) return;
if (!isAuthenticated) {
if (redirectRef.current) return;
redirectRef.current = true;
void navigate({
to: '/login',
search: { redirect: redirectTarget },
replace: true,
}).catch(() => {
redirectRef.current = false;
});
} else {
redirectRef.current = false;
}
}, [isAuthenticated, isPending, navigate, redirectTarget]);
if (isPending || !isAuthenticated) {
return <ShellSkeleton />;
}
return (
<AppChrome>
<Outlet />
</AppChrome>
);
}
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 First seen · 277 lines · 2,686 tokens per session scan A 1d4eb2cff2ee
tanstack-start-template AGENTS.md is an instructions file published in the GitHub repository dyeoman2/tanstack-start-template (26 stars, last pushed 4mo ago), licensed MIT. It adds 2,686 tokens to every session, about $0.0134 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 2026-09-01.
Other instructions, from other repositories
modernstack-saas AGENTS.md
Instructions for joachimchauvet/modernstack-saas, covering project guidelines, core technologies, key conventions, svelte best practices and high-level principles.
vexpo AGENTS.md
Instructions for ramonclaudio/vexpo, covering agents.md, project, stack rules, monorepo (root) and template (templates/default/).
modernstack-saas CLAUDE.md
Instructions for joachimchauvet/modernstack-saas, a project described as: A modern SaaS template/boilerplate built with SvelteKit, Convex, Better Auth, and shadcn-svelte. Includes subscriptions, auth, user & admin dashboard, user settings, and more.
goai AGENTS.md
Instructions for zendev-sh/goai, covering agents.md - goai, commands, architecture, key rules and adding providers.
ai-sdk-provider-claude-code AGENTS.md
Instructions for ben-vargas/ai-sdk-provider-claude-code, covering repository guidelines, project structure & module organization, build, test, and development commands, coding style & naming conventions and testing guidelines.
flarestarter CLAUDE.md
Instructions for FlareStarter/flarestarter: This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.