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.
git clone --depth 1 https://github.com/ammonhaggerty/a-primer-skillsWrote 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.
[](https://agentmods.dev/commands/ammonhaggerty/a-primer-skills/a-new-project)<a href="https://agentmods.dev/commands/ammonhaggerty/a-primer-skills/a-new-project"><img src="https://agentmods.dev/badge/commands/ammonhaggerty/a-primer-skills/a-new-project/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/ammonhaggerty/a-primer-skills/a-new-project"><img src="https://agentmods.dev/badge/commands/ammonhaggerty/a-primer-skills/a-new-project.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00034 | $0.05469 |
| Opus 5 | $0.00017 | $0.02735 |
| Sonnet 5 | $0.00007 | $0.01094 |
| Haiku 4.5 | $0.00003 | $0.00547 |
Grade A, and why
a-new-project 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 10d 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.
How it starts
The opening of the file, as written. The whole thing — 658 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/a-new-project
Scaffold a new full-stack project. This command is repeatable — run it every time a new project is needed.
Follow all seven steps in order. Complete each step before moving to the next.
Step 1: Workspace Check
Detect the operating system and check the current working directory.
- Mac/Linux: Check if the current directory is
~/Development. - Windows: Check if the current directory is
%USERPROFILE%\Development.
If not in the workspace directory:
- Explain the convention: "Keeping all projects in one place makes them easy to find and gives Claude consistent workspace awareness."
- Check whether
~/Development(or the Windows equivalent) exists. If it does not, offer to create it. - Navigate to the workspace directory before proceeding.
Step 2: Project Choice
Choose a starting point
Ask:
How would you like to start?
Sample promo page — A working single-page personal site with profile card, weather, social links, AI chat, timeline, and music widget. You'll customize it with your info, then modify or add modules from there.
Empty project — A clean slate with the full stack wired up (Hono, DaisyUI, HTMX, Alpine.js, D1, R2, Workers AI) but no modules. Good for building something completely new.
Store the choice as PROJECT_MODE (sample or empty).
Project name
If sample mode:
Ask:
What should we call your project? This becomes the folder name and URL. A short kebab-case name works best — for example,
my-site,dj-ammon,jane-codes.
If empty mode:
Ask:
What do you want to build? Describe it in a few sentences — who it's for, what it does, what the core experience feels like.
Then ask:
What should the project be called?
Suggest a kebab-case name derived from their description (e.g., daily-checkin, recipe-finder). Let the user accept or choose a different name.
Store the final name as PROJECT_NAME.
Step 3: Scaffold
Create the project directory at ~/Development/PROJECT_NAME/.
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.
- 10d ago First seen · 658 lines · 34 tokens per session scan A a1193c003d3b
a-new-project is a command published in the GitHub repository ammonhaggerty/a-primer-skills (3 stars, last pushed 1mo ago), licensed MIT. It adds 34 tokens to every session and 5,469 once invoked, about $0.0002 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.
Other commands, from other repositories
matilha-install
Interactive wizard to install the matilha ecosystem — preset packs + optional CLAUDE.md bootstrap. Falls back to static guide if wizard tools are unavailable.
ml-impl
Full implementation loop - plan, write, and review code (Meta-Loop Steps 3.1-3.4).
ml-wireframe
Review wireframe to ensure AI understanding before spec creation (Meta-Loop Step 1.2).
ml-retain
Capture session learnings in documentation (Meta-Loop Step 4).
gather
Phase 40 — merge completed SPs in wave order, run per-SP regression, update wave-status.
plan
Phase 20-30 — author spec + plan with waves, SPs, mergeorder, exit gates.