scaffold AGENTS.md

Project instructions for scaffold, a standard for starting or organising Howells projects, including repository layout, technology choices, package boundaries, and release checks.

In plain words
What is it for?
Use them when creating or standardising a project, deciding where applications and packages belong, choosing the default web stack, and following the repository’s development and deployment rules.
Why use it?
They give a project a consistent structure and define which files and generated outputs are the source of truth.

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

Made for: Codex, OpenCode.

Per session 711 This file is loaded in full into every session.
When invoked 711 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 $0.00711 $0.00711
Opus 5 $0.00356 $0.00356
Sonnet 5 $0.00142 $0.00142
Haiku 4.5 $0.00071 $0.00071

Measured 2d ago against content hash e0d33d05d0c4, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

scaffold 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 · 44 lines

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.

Scaffold

The house baseline for starting or standardising a Howells project. It defines repo shape, stack, package boundaries, agent workflow, deployment, and launch checks. It is a standard, not a template.

docs/ is the source of truth. skills/scaffold is a distribution surface, generated from it.

What a new project takes from here

Read docs/README.md first, then open only the reference page needed for the task.

Defaults a new repo inherits:

  • pnpm monorepo, usually apps/* and packages/*, with Turborepo and cache disabled until the repo proves it's deterministic.
  • Next.js App Router, React, Tailwind v4, Base UI, and Storybook when the repo exports reusable UI; unified Radix is the deliberate opt-out.
  • @howells/lint on the Oxlint/Oxfmt lane, @howells/typescript-config for tsconfig presets, @howells/envy for typed env parsing and deploy-time env checks.
  • Drizzle and Neon for product data, the narrowest typed API boundary that fits, and React Query when the client owns server-state.
  • @howells/ai as the provider baseline, Motif's SDK or agent-readable CLI for image work, product orchestration in a repo-local ai or agents package, and MCP in its own package only when the product genuinely exposes it.

docs/reference/agent-workflow.md defines what a root AGENTS.md should cover. Other decisions live in docs/reference/; deviations live in docs/adr/.

Generated surfaces - never hand-edit

  • docs/principles.md is generated from src/content/principles.ts by pnpm generate:principles. The homepage renders the same module, so they can't drift.
  • skills/scaffold/references/ is generated from docs/ by pnpm sync:skill. Edit docs once, then run the sync.
  • pnpm check runs both with --check and fails on drift, so a docs edit without a regenerate breaks the gate.

Editing this repo

  • Don't turn scaffold docs into guidance for one repo. If it only applies to one project, it belongs in that project's AGENTS.md.
  • Don't vendor independent skills into product repos; link or install them.
  • Search docs/reference/ before adding new baseline policy, and search existing pages before creating a new one.
  • Search CONTEXT.md before changing the terms Scaffold Baseline, Project Docs, Agent Skill, Agent Skill Distribution or Skill Wrapper.
  • Don't add runtime dependencies unless the docs or skill distribution genuinely needs them.
  • Write short, declarative prose. Cut scene-setting, recaps, fake contrasts, vague praise, and claims without evidence.
  • Prefer a paragraph or short list to a table unless exact comparison is the point.

Read the full file on GitHub · 44 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 · 44 lines · 711 tokens per session scan A e0d33d05d0c4

Subscribe to this mod's changes

scaffold AGENTS.md is an instructions file published in the GitHub repository howells/scaffold (2 stars, last pushed 7d ago), licensed MIT. It adds 711 tokens to every session, about $0.0036 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 2026-08-31.