agent-building

A guide for designing and creating packaged AI agents, including their identity, instructions, capabilities, files, and visual assets.

In plain words
What is it for?
Use it to create an agent manifest, role instructions, policies, context files, README documentation, and avatar or screenshot assets.
Why use it?
It gives the required structure and file locations for agents that can be installed and recognised by the system.

Skill for Claude CodeCodex

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 skills/markus-global/markus/agent-building
Any agent
npx skills add markus-global/markus --skill agent-building
Clone the repo
git clone --depth 1 https://github.com/markus-global/markus

Made for: Claude Code, Codex.

Per session 33 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 4,088 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 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 $0.00033 $0.04088
Opus 5 $0.00016 $0.02044
Sonnet 5 $0.00007 $0.00818
Haiku 4.5 $0.00003 $0.00409

Measured 3d ago against content hash 1e82a3b0dc7b, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

agent-building scanned grade C 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 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.

Recursive force deletehighDestructive command

rm -rf with a variable or a broad path is one typo away from removing the wrong tree.

- "Never run `rm -rf` or other destructive commands"
templates/skills/agent-building/SKILL.md · 279 lines

How it starts

The opening of the file, as written. The whole thing — 279 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agent Building

This skill teaches you how to create Markus agent packages — self-contained directory-based artifacts that define an AI agent's identity, capabilities, constraints, and visual assets.

Artifact Directory

CRITICAL: Agent artifacts MUST be saved under this exact path — the Builder page, install system, and deliverable detection all depend on it:

~/.markus/builder-artifacts/agents/{agent-name}/
├── agent.json       # Manifest (auto-created from your JSON output)
├── README.md        # Public-facing overview for Hub/Builder (REQUIRED)
├── ROLE.md          # Identity and system prompt (REQUIRED)
├── HEARTBEAT.md     # Periodic self-check checklist (RECOMMENDED)
├── POLICIES.md      # Constraints & guardrails (optional)
├── CONTEXT.md       # Domain context & references (optional)
└── images/          # Image assets (avatar, screenshots)
    └── avatar.jpg   # Agent avatar/thumbnail image

Do NOT write artifacts to ~/.markus/shared/, your working directory, or any other location. Only ~/.markus/builder-artifacts/agents/ is recognized by the system.

When the user installs the artifact, files are deployed to ~/.markus/agents/{agentId}/role/. The ROLE.md becomes the agent's system prompt — it IS the agent's identity, not an override of a template.

Package Slug (name) — REQUIRED

The manifest name is the package slug: directory name, Hub URL segment (/@user/{slug}), and share/publish id.

Rules (hard — invalid manifests are rejected on write / save / share):

  • English kebab-case only: lowercase letters a-z, digits 0-9, hyphens -
  • 2–64 characters; must start with a letter
  • Pattern examples: code-reviewer, paper-mentor, seo-auditor
  • NOT allowed: Chinese (智库研究团队), spaces, underscores, UPPERCASE, emoji, or empty
  • Put the human-readable title (any language) in displayName, never in name
User language name (slug) displayName
Chinese "论文导师" paper-mentor 论文导师
English "Code Reviewer" code-reviewer Code Reviewer

Read the full file on GitHub · 279 lines

Files

What ships with it

1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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. 3d ago First seen · 279 lines · 33 tokens per session scan C 1e82a3b0dc7b

Subscribe to this mod's changes

agent-building is a skill published in the GitHub repository markus-global/markus (157 stars, last pushed 6d ago), licensed Apache-2.0. It adds 33 tokens to every session and 4,088 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.