workshops

Repository rules for small coding experiments, called workshops. They define where experiments, longer-lived projects, paused work, references, and retired experiments belong.

In plain words
What is it for?
Use them when starting an experiment, promoting it into a longer-lived project, pausing work, storing references, or documenting meaningful changes.
Why use it?
They give new work a predictable location and require focused edits, relevant documentation updates, and records for substantial decisions or progress.

Cursor rule for Cursor

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 rules/andyuneducated/alchemy/workshops
Clone the repo
git clone --depth 1 https://github.com/AndyUneducated/alchemy

Made for: Cursor.

Per session 834 This file is loaded in full into every session.
When invoked 834 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 $0.00834 $0.00834
Opus 5 $0.00417 $0.00417
Sonnet 5 $0.00167 $0.00167
Haiku 4.5 $0.00083 $0.00083

Measured yesterday against content hash 48ef7c4ebfb3, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

workshops 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 yesterday.

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.

.cursor/rules/workshops.mdc · 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.

Workshops (vibe coding)

Repo layout

Path Use
play/ Default for new spikes, scripts, disposable demos
grow/ Longer-lived mini-apps promoted from play/
stash/ Paused WIP
refs/ Copied reference snippets, not first-class product code
_archive/ Retired experiments

Unless the user names another path, create new experiments under play/.

Editing discipline

  • Minimal diffs: only what the task needs; no drive-by refactors, unrelated files, or rewrites of untouched sections.
  • Docs: after a change, check existing docs (README, DECISIONS, JOURNAL, etc.) for items that need updating and update them in the same turn; only add new entries (sections, list items, new doc files) when the user asks. Two exceptions where appending is mandatory rather than gated on user request: a non-trivial technical decision always warrants a new ADR entry in DECISIONS.md, and a working day's substantive progress always warrants ≤2 milestone entries in JOURNAL.md (see "Decision & journal logs" below).
  • Doc formatting: default to tables for any content with 2+ parallel items (enumerations, lists, comparisons, option sets); use bullets/lists only when items aren't parallel or share no dimensions.
  • Markdown: CommonMark / GFM conventions; compact tables (minimum separator dashes, no cell padding).
  • No new tooling: don't add lint/test/build infrastructure where none exists; if checks are already defined, run them after non-trivial edits.
  • Commits: split by independent semantic increments, but never at the cost of an unusable intermediate state; commit messages must be in English only.

Decision & journal logs

Each play/ (or grow/) sub-project with non-trivial design choices keeps two append-only logs alongside the README:

File Unit Records
DECISIONS.md one architecture decision Status + Date metadata, then Context / Options considered / Decision / Consequences (ADR-style, MADR-flavored). Sub-projects may keep their own section headings (e.g. Scope / Implementation for play/evals phase entries) as long as the metadata lines are present.
JOURNAL.md one milestone (≤2 per working day) ## YYYY-MM-DD — title heading; 功能 / 技术 sections required; 取舍 section optional and should reference DECISIONS §N instead of duplicating rationale.

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. yesterday First seen · 48 lines · 834 tokens per session scan A 48ef7c4ebfb3

Subscribe to this mod's changes

workshops is a cursor rule published in the GitHub repository AndyUneducated/alchemy (20 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 834 tokens to every session, about $0.0042 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.