ship-it AGENTS.md

Project instructions for the ship-it coding-agent skill, including its repository layout, terminology, and component structure.

In plain words
What is it for?
Use them when changing the ship-it skill, which checks common product and user-experience details such as login, checkout, billing, and error pages.
Why use it?
They give the agent the context needed to navigate the project and follow its conventions consistently.

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/lunkibr/ship-it/agents-md
Clone the repo
git clone --depth 1 https://github.com/LunkiBR/ship-it

Made for: Codex, OpenCode.

Per session 1,343 This file is loaded in full into every session.
When invoked 1,343 The same file — it is already loaded in full.
Security scan A 0 findings. 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 $0.01343 $0.01343
Opus 5 $0.00672 $0.00672
Sonnet 5 $0.00269 $0.00269
Haiku 4.5 $0.00134 $0.00134

Measured 2d ago against content hash 14cc486eda5e, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

ship-it 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 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.

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.

AGENTS.md · 48 lines

How it starts

The opening of the file, as written. The whole thing — 48 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

Project Context

This repository is the source for a single Claude Code Agent Skill (ship-it) that helps an AI coding agent catch commonly-forgotten UX/product details on a screen or feature before it ships: login, checkout, billing, pricing, 404s, admin panels, and around 60 other well-known patterns. It packages a large, still-growing catalog as a router (SKILL.md) → section indexes (sections/) → per-pattern reference files (references/), so an agent loads only what a specific task needs rather than the whole catalog every time.

Repository Layout

  • SKILL.md — the skill's entry point: frontmatter (name, description) plus workflow steps and the judgment/tier rules. Read this first; it never contains pattern content, only pointers to it.
  • sections/*.md — one flat index per section, listing that section's patterns (name, tier, one-liner, link to its reference file). Nothing here is opened unless SKILL.md has already narrowed the task to that section.
  • references/*.md — one file per pattern, the actual checklist. TEMPLATE.md is the exact format every new one follows; billing.md and login.md are the finished quality bar.
  • generation/ — not part of the skill itself. PROMPT.md is the brief used to bulk-author reference files through a cheaper model, grounded in established UX patterns, product standards, and primary references rather than generated from model intuition alone.
  • scripts/validate.py — the one automated check this project has: SKILL.md's frontmatter and line budget against the Agent Skills spec, and every reference file's format against this project's own conventions. Run before committing changes to SKILL.md or references/.
  • ARCHITECTURE.md — why the project is shaped this way: one skill instead of many, the tier system, the naming decisions (and the one that got reverted), which real precedents were studied.
  • README.md — the human-facing pitch and current build status.

Terminology

  • Section — one of the four top-level surfaces: Sidekick (mobile app), Control Room (in-product web app), Storefront (marketing website), Choreography (cross-screen interaction flows). These four names are deliberately original branding — see ARCHITECTURE.md's naming section for why, and why that's the only place originality was worth the cost.
  • Pattern — one screen or interaction moment (Login, Cart, 404, Making a card payment...). Always the industry-standard name. Never renamed, regardless of how tempting — see the reverted attempt logged in ARCHITECTURE.md.
  • TierFundamental, Common, or Conditional. The default assumption for whether a pattern applies to a given product, fixed once per pattern so an agent's "use judgment" has an actual anchor instead of drifting toward whichever answer is less work.
  • Cross-cutting pattern — a pattern that appears in more than one section's index (Billing, Login, Account, Settings, Search, Cart) but still resolves to exactly one reference file, written once.

Read the full file on GitHub · 48 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 · 48 lines · 1,343 tokens per session scan A 14cc486eda5e

Subscribe to this mod's changes

ship-it AGENTS.md is an instructions file published in the GitHub repository LunkiBR/ship-it (18 stars, last pushed 23d ago), licensed MIT. It adds 1,343 tokens to every session, about $0.0067 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.