mikropkl AGENTS.md

mikropkl AGENTS.md is an instructions file for Codex, OpenCode from tikoci/mikropkl. It costs 7,381 tokens per session, scanned A, original, Unlicense.

Project instructions for coding agents working on mikropkl, a tool that creates virtual-machine packages for MikroTik RouterOS using declarative configuration files.

In plain words
What is it for?
Use them when modifying or troubleshooting UTM bundles, QEMU launch files, virtual disks, or related build and networking behavior.
Why use it?
They explain the project's architecture, image pipeline, platform differences, and important implementation decisions so agents can make informed changes.

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/tikoci/mikropkl/agents-md
Clone the repo
git clone --depth 1 https://github.com/tikoci/mikropkl

Made for: Codex, OpenCode.

Wrote 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.

agentmods badge for mikropkl AGENTS.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/tikoci/mikropkl/agents-md.svg)](https://agentmods.dev/instructions/tikoci/mikropkl/agents-md)
Your own site
<a href="https://agentmods.dev/instructions/tikoci/mikropkl/agents-md"><img src="https://agentmods.dev/badge/instructions/tikoci/mikropkl/agents-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 7,381 This file is loaded in full into every session.
When invoked 7,381 The same file — it is already loaded in full.
Security scan A 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 $0.07381 $0.07381
Opus 5 $0.03691 $0.03691
Sonnet 5 $0.01476 $0.01476
Haiku 4.5 $0.00738 $0.00738

Measured 4d ago against content hash 3a65b107a6fb, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

mikropkl AGENTS.md scanned grade A 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 4d 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl -s -m 5 http://localhost:9180/ | head -1
AGENTS.md · 509 lines

How it starts

The opening of the file, as written. The whole thing — 509 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md — Agent Reference: Roadmap, Implementation Notes, and Decision Log

This document captures deeper technical context, architecture decisions, and planned work items for AI coding agents working on this project. It supplements CLAUDE.md (which is the primary project reference) with information useful for multi-session agent work.

Architecture Summary

The project has two output layers, each derived from the same pkl manifests:

   pkl Manifests (.pkl)
        │
        ├─→ UTM bundle (.utm)       macOS/UTM.app — primary user-facing
        │     config.plist
        │     Data/ (disk images)
        │     qemu.cfg               QEMU --readconfig ini format
        │     qemu.sh                Launch script (pflash, accel, networking, serial)
        │
        └─→ (libvirt.xml)           Experimental — disabled by default

The .utm directory IS a ZIP archive — this is the key insight for Linux deployment. UTM bundles are macOS-first but structurally portable. Every bundle includes qemu.cfg + qemu.sh so users on Linux (or macOS without UTM) can unpack the ZIP and run QEMU directly.

Image Pipeline

MikroTik download   ─→  chr-<ver>.img          (x86_64, proprietary boot)
                    ─→  chr-<ver>-arm64.img    (aarch64, standard FAT16 EFI)

tikoci/fat-chr      ─→  chr-efi.img            (x86_64, reformatted FAT16 EFI)
                        (required for Apple VZ / UEFI boot)

When to use which image:

  • QEMU x86_64 → standard chr-<ver>.img + SeaBIOS (simplest, fastest)
  • QEMU aarch64 → standard chr-<ver>-arm64.img + EDK2 UEFI pflash
  • Apple VZ x86_64 → chr-efi.img from fat-chr (needs proper FAT EFI partition)
  • Apple VZ aarch64 → [untested] standard ARM64 image (already FAT16)

VirtIO Deep Dive

This is the single most important QEMU detail for this project. Getting the disk interface wrong is the #1 cause of boot failures.

Why UTM says "NVMe" but uses virtio-blk-pci on aarch64

UTM's translation layer maps plist Interface=NVMe differently per backend:

  • Apple backend: actual NVMe (Virtualization.framework handles it)
  • QEMU backend: UTM passes -device virtio-blk-pci regardless of the plist value

Read the full file on GitHub · 509 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. 4d ago First seen · 509 lines · 7,381 tokens per session scan A 3a65b107a6fb

Subscribe to this mod's changes

mikropkl AGENTS.md is an instructions file published in the GitHub repository tikoci/mikropkl (52 stars, last pushed today), licensed Unlicense. It adds 7,381 tokens to every session, about $0.0369 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.