artifacts CLAUDE.md

A set of project instructions for a self-hosted service that publishes HTML, React components, Markdown, or static-site archives at private URLs. Self-hosted means the service runs on infrastructure you control.

In plain words
What is it for?
It is for guiding work on the service, its domains, documentation, tests, storage backends, and artifact-publishing behavior.
Why use it?
It gives coding agents the product rules they must follow, including the renamed domains, storage layout, testing commands, and documentation requirements.

Instructions file

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/anvilnine/artifacts/claude-md
Clone the repo
git clone --depth 1 https://github.com/anvilnine/artifacts
Per session 497 This file is loaded in full into every session.
When invoked 497 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 $0.00497 $0.00497
Opus 5 $0.00249 $0.00249
Sonnet 5 $0.00099 $0.00099
Haiku 4.5 $0.00050 $0.00050

Measured 2d ago against content hash 951c5cf02260, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

artifacts CLAUDE.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.

CLAUDE.md · 37 lines

What it actually says

artifacts

Self-hosted, Claude-style artifact publishing. POST HTML / a React component / Markdown / a zipped static site, get back an unguessable URL on your own domain. One container, single admin account, no database by default (plain files under /data); optional S3 / git / Postgres / SQLite backends. No build step, Node >= 22. The end-to-end suite is bash .github/workflows/smoke.sh <url> <key>; unit tests run with npm test.

Read the README and docs/ before working on a feature.

Domains

This product is being renamed to Dropkiln (docs/superpowers/plans/backlogs/backlog-6-rebrand.md). Both domains are bought and both now resolve through Cloudflare nameservers on the Anvil Nine Cloudflare account, as of 2026-08-06:

  • dropkiln.com = company and product: marketing, docs, blog, dashboard at app.dropkiln.com, email. Cookies live here.
  • dropkiln.app = user content only: tenant sites and artifacts at <name>.dropkiln.app. Wildcard, Cloudflare for SaaS custom hostnames, and the routing Worker all attach to this zone. Never set a dashboard cookie here.

Backlog 6 T6.1.1 said anvilnine.com would stay off Cloudflare. That is no longer true; all five Anvil Nine zones are on the one account now. Nothing about the .com / .app split changes.

Current state worth knowing before you plan anything DNS-shaped: both apexes have proxied A records pointing at an origin that is down, so https://dropkiln.com returns 522. Neither zone has a DMARC record, so the T6.1.2 parking task (null MX + SPF + DMARC on both) is still open. Org-level table: ../docs/domains.md.

Writing docs (README, docs/, PR bodies, comments)

Follow the org-wide prose rules in the parent anvilnine/CLAUDE.md.

Repo-specific check: before committing a README change, run grep -n "—\|–" README.md and confirm it returns nothing.

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 · 37 lines · 497 tokens per session scan A 951c5cf02260

Subscribe to this mod's changes

artifacts CLAUDE.md is an instructions file published in the GitHub repository anvilnine/artifacts (7 stars, last pushed 3d ago), licensed MIT. It adds 497 tokens to every session, about $0.0025 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.