agents-md AGENTS.md

Project instructions for building Convex backends with Svelte or SvelteKit frontends. Convex is a backend platform, while Svelte and SvelteKit are tools for building web interfaces and applications.

In plain words
What is it for?
Use it when developing Convex functions or Svelte/SvelteKit interfaces in the same project, especially when adding data validation, UI components, or backend integration.
Why use it?
It sets shared rules for connecting the frontend to the backend, validating backend inputs, using TypeScript, formatting code, and checking the project before commits.

Instructions file for CodexOpenCode

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 instructions/ytgz/agents-md/agents-md
Clone the repo
git clone --depth 1 https://github.com/YtGz/agents-md

Made for: Codex, OpenCode.

Per session 14,103 This file is loaded in full into every session.
When invoked 14,103 The same file — it is already loaded in full.
Security scan B 1 finding. 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 $0.14103 $0.14103
Opus 5 $0.07051 $0.07051
Sonnet 5 $0.02821 $0.02821
Haiku 4.5 $0.01410 $0.01410

Measured 2d ago against content hash 128d9c38e673, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

agents-md AGENTS.md scanned grade B with 1 finding 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 2d 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.

Sends data to an external URLmediumData exfiltration

A POST to an outside endpoint may be telemetry or may be exfiltration; either way the mod talks to somewhere, and you should know where.

const response = await fetch("https://api.example.com", { method: "POST",
AGENTS.md · 2,019 lines

How it starts

The opening of the file, as written. The whole thing — 2,019 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agent Instructions for Convex with Svelte

This is your comprehensive guide for working with Convex in Svelte/SvelteKit projects.

Overview

You have access to three powerful MCP servers:

  1. Convex MCP Server - For interacting with Convex deployments
  2. Svelte MCP Server - For accessing Svelte 5 and SvelteKit documentation
  3. GitHub MCP Server - For repository management, issues, pull requests, and more

You also have access to:

  • Bits UI Documentation - Headless component library docs available via fetch (llms.txt standard)

Priority Rules

CRITICAL: When working with Convex and Svelte together:

  1. Always prefer convex-svelte patterns for frontend integration
  2. Always use Convex MCP tools over command-line tools where possible
  3. Follow Svelte 5 runes and reactive patterns when writing frontend code
  4. Use proper validators and TypeScript types in Convex backend functions
  5. Use Bits UI for headless components - fetch docs on demand when building UI
  6. Always use Conventional Commits for all git commits
  7. Always run bun run format and bun run check before committing
  8. Always use the $convex alias when importing from Convex in Svelte/SvelteKit files (e.g., import { api } from '$convex/_generated/api')

Project Bootstrap

IMPORTANT: You MUST complete ALL bootstrapping steps before creating any application-specific code. The initial commit contains ONLY the project skeleton/boilerplate. Application-specific code is written AFTER the user has started the dev servers.

When starting a new project, you will typically be in an empty folder containing only an AGENTS.md file (and possibly a LICENSE.md file). Follow this workflow to bootstrap the project:

Step 1: Initialize SvelteKit Project

Run the following command to scaffold a SvelteKit app with our standard configuration:

bunx sv create . --template minimal --types ts --add eslint --add prettier --add vitest="usages:unit,component" --add tailwindcss="plugins:typography,forms" --add paraglide="languageTags:en,de+demo:no" --no-dir-check --install bun

Read the full file on GitHub · 2,019 lines

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. 2d ago First seen · 2,019 lines · 14,103 tokens per session scan B 128d9c38e673

Subscribe to this mod's changes

agents-md AGENTS.md is an instructions file published in the GitHub repository YtGz/agents-md (2 stars, last pushed 4mo ago), licensed MIT. It adds 14,103 tokens to every session, about $0.0705 per session on Opus 5. A static security scan graded it B with 1 finding (sends data to an external url). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.