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/karngyan/karngyan.com/claude-mdgit clone --depth 1 https://github.com/karngyan/karngyan.comWrote 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/instructions/karngyan/karngyan.com/claude-md)<a href="https://agentmods.dev/instructions/karngyan/karngyan.com/claude-md"><img src="https://agentmods.dev/badge/instructions/karngyan/karngyan.com/claude-md.svg" alt="Measured on agentmods" 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 | $0.00817 | $0.00817 |
| Opus 5 | $0.00409 | $0.00409 |
| Sonnet 5 | $0.00163 | $0.00163 |
| Haiku 4.5 | $0.00082 | $0.00082 |
Grade A, and why
karngyan.com CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 44 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
Personal-site template: TanStack Start + React 19 + Tailwind CSS v4 + shadcn/ui + MDX, deployed to Cloudflare Workers. Forkers personalize via site.config.ts and src/content/.
Commands
mise install # pins node + pnpm (see mise.toml)
pnpm install
pnpm dev # dev server on http://localhost:3000
pnpm build # generates OG images, then vite build → .output/
pnpm test # vitest
pnpm check # prettier --write + eslint --fix
pnpm typecheck # tsc --noEmit
pnpm generate:og # regenerate OG images only (satori → sharp)
pnpm run deploy # build + wrangler deploy (Cloudflare Workers)
pnpm changeset # record a change (patch/minor/major + summary)
pnpm changeset:version # consume changesets -> bump version + CHANGELOG.md
Architecture
site.config.ts— single typed config for identity, SEO, socials, nav, resume toggle, work history, analytics. Import it via a relative path (../../site.config); the@/alias only coverssrc/.src/routes/— TanStack Router file routes. Pages areindex.tsx/$slug.tsx; server routes use theserver.handlers.GETpattern with bracket-escaped filenames (rss[.]xml.ts,sitemap[.]xml.ts,robots[.]txt.ts,llms[.]txt.ts).routeTree.gen.tsis generated — never edit.src/content/articles/<slug>/page.mdx— articles. Each exportsconst article = { title, description, date, author }and default-exportsArticleLayout.src/lib/articles.tseager-globs these at build time; RSS/sitemap/OG all derive from it.src/lib/feeds.ts— RSS + robots.txt builders (unit-tested infeeds.test.ts).src/components/— layout components at top level, shadcn primitives inui/(add more withpnpm dlx shadcn@latest add <name>).src/styles.css— Tailwind v4 CSS-first config: design tokens as CSS variables (oklch), dark mode via.darkclass +@custom-variant. No tailwind.config.js.scripts/generate-og-images.tsx— build-time OG images; reads article metadata with a regex (article exports must stay statically analyzable).wrangler.jsonc— Cloudflare Workers config; build output lands in.output/(server entry + public assets).
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.
- 6d ago First seen · 44 lines · 817 tokens per session scan A 41f7336bd76e
karngyan.com CLAUDE.md is an instructions file published in the GitHub repository karngyan/karngyan.com (369 stars, last pushed 1mo ago), licensed MIT. It adds 817 tokens to every session, about $0.0041 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
gitprofile AGENTS.md
AGENTS.md instructions for arifszn/gitprofile, covering agents.md, what this project is, customizing a portfolio: edit one file, the base option is the most common source of breakage and layout.
gitprofile CLAUDE.md
Claude Code instructions for arifszn/gitprofile, a project described as: 🚀 Create and publish a GitHub-synced portfolio by just providing your GitHub username.
easonchang.com-next AGENTS.md
AGENTS.md instructions for eason-dev/easonchang.com-next, covering agents.md — guide for ai coding agents, commands, architecture and conventions.
easonchang.com-next CLAUDE.md
Claude Code instructions for eason-dev/easonchang.com-next, covering claude-specific guidance, skills, hard rules and definition of done.
personal-site AGENTS.md
Instructions for mldangelo/personal-site, covering agents.md, quick commands, project structure, code style and git workflow.
devportfolio CLAUDE.md
Instructions for RyanFitzgerald/devportfolio, covering claude.md, project overview, tech stack, development commands and architecture.