mad-science-setup

Setup Mad Science Skills

1. Explore Look at the current repo to understand its starting state. Read whatever exists; don't assume:

• git remote -v and .git/config: is this a GitHub repo? Which one?
• AGENTS.md and CLAUDE.md at the repo root: does either exist? Is there already an ## Agent skills section in either?
• GLOSSARY.md and GLOSSARY-MAP.md at the repo root
• Monorepo signals: a pnpm-workspace.yaml, a workspaces field in package.json, or a populated packages/* with its own src/. These are present only in a genuinely large multi-package repo; their absence means single-context, which is almost every repo.

2. Show the Finds and Ask

Summarize what's present and what's missing. Then take the sections in order. One section, one answer, then the next.

Section A:

Experiment tracker: The experiment tracker lives in this repo at docs/experiments-pipeline.

Section B:

Domain docs: Default to single-context (one GLOSSARY.md at the repo root). This fits almost every repo; write it without asking.

Offer multi-context (a root GLOSSARY-MAP.md pointing to per-context GLOSSARY.md files) only when exploration found monorepo signals. Then confirm which layout they want.

3. Confirm and edit

Show the user a draft of:

• The ## Agent skills block to add to whichever of CLAUDE.md / AGENTS.md is being edited (see step 4 for selection rules)

4. Write Pick the file to edit:

• If CLAUDE.md exists, edit it.
• Else if AGENTS.md exists, edit it.
• If neither exists, ask the user which one to create; don't pick for them.

Never create AGENTS.md when CLAUDE.md already exists (or vice versa); always edit the one that's already there.

If you create the file, add these rules to it:

• The file must indicate that the docs/ folder is the main source of truth for the experiments.
• Add a rule to keep this file in sync with the docs folder, updated as experiments are created.
• Create a section in the file with this template:

## Docs Indice


### Articles
<article_name_file.md> : <article short description>

### Experiments Docs
<experiments_name_file.md> : <experiment short summary> - 
** Already complete? ** : Answer yes or no. If yes, add the date.


### Experiments Pipeline
<experiments_pipeline_name_file.md> : <experiment pipeline short summary> - 
** Already complete? ** : Answer yes or no. If yes, add the date.

5. Optional — Configure "notify human wait"

Ask the user whether they want the repo to notify a human when a flow skill is in progress and the turn waits for human input. Explain briefly what it does (Google Chat card with the open question + project name). One yes/no.

• If no: skip the section, continue.
• If yes:
  • Add GCHAT_WEBHOOK_URL=CHANGE_ME to .env.example (if not present) and tell the user to set the real webhook URL in .env (never versioned).
  • Add the "Flow skills — notify human wait" block to CLAUDE.md (or the file being edited), listing the flow skills currently in this repo (e.g. let-it-rip, literature-research, to-article) plus the rule "notify whenever work inside a listed flow skill is waiting for a human response". Keep the list maintainable: new flow skills are added here.

6. Done

Tell the user the setup is complete and which engineering skills will now read from these files.