substances — Claude Code skill: quitting plans and installation
The substances Claude Code skill builds plans to quit or reduce smoking, vaping, or alcohol use. It checks for risks from stopping without supervision and sets a stepwise schedule.
MIT ★ 93 updated 2026-11-15
npx skills add Insilico-org/plsi --skill substances -a claude-codeWhat substances does
The skill gathers details about substance use, earlier quit attempts, and triggers, then proposes a quit or reduction date with a gradual schedule. For alcohol, it also asks about withdrawal symptoms. Heavy daily drinking or any withdrawal sign makes a clinician visit the first step; a history of withdrawal seizures or confusion is flagged as urgent.
Schedules and support options draw on verified PubMed evidence. The skill predicts changes against recorded cigarette or alcohol amounts, then sends its proposal to PLSI Nostradamus in work/<thread>/proposal-coach.json. It does not prescribe medication or check and adopt the plan itself; medication options go through a separate review.
What it helps with
- plan a gradual smoking quit schedule
- reduce alcohol use in steps
- check withdrawal risks before stopping abruptly
- prepare responses to common triggers
Example prompts for the agent
- Help me build a gradual plan to quit smoking.
- I want to drink less; assess withdrawal risks and make a reduction schedule.
- I started smoking again after a quit attempt; account for that in a new plan.
How to install substances
1Skills CLI (Vercel)
npx skills add Insilico-org/plsi --skill substances -a claude-code2Manually: put the files in place
~/.claude/skills/substances/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/substances/
1Skills CLI (Vercel)
npx skills add Insilico-org/plsi --skill substances -a codex2Manually: put the files in place
.agents/skills/substances/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add Insilico-org/plsi --skill substances -a cursor2Manually: put the files in place
.cursor/skills/substances/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install Insilico-org/plsi skills/coach/substances2Manually: put the files in place
.github/skills/substances/SKILL.md1Skills CLI (Vercel)
npx skills add Insilico-org/plsi --skill substances -a gemini-cli2Platform CLI
gemini skills install https://github.com/Insilico-org/plsi --path skills/coach/substancesManually: put the files in place
.devin/skills/substances/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about substances
How do I install substances in Claude Code?
Run npx skills add Insilico-org/plsi --skill substances -a claude-code. The skill file is at ~/.claude/skills/substances/SKILL.md.
What does substances do?
It helps create a plan to quit or reduce smoking, vaping, or alcohol use. The skill checks risks from unsupervised withdrawal, builds a stepwise schedule, and sends the proposal to PLSI Nostradamus.
What does substances need to work?
It needs PLSI to read context, search and verify PubMed evidence, and pass the proposal to PLSI Nostradamus.
Source text
SKILL.md · MIT
--- name: substances description: Use as PLSI Coach when the user wants to quit or cut down smoking, vaping or alcohol - check for risks of an unsupervised stop, set a quit or reduce date with a stepwise schedule from run-time evidence, and predict on the coordinates that exist. --- # PLSI smoking and alcohol plan ## When to use - "Help me quit smoking", "I want to drink less", a goal in `personal_context`, or PLSI routes it here. - Not under a `scope:` line (`orchestrator/urgent`) that covers it (pregnancy: point to the obstetric clinician, who can still help with quitting). ## Steps 1. **Read the state:** ```bash plsi state show --instance <instance> --map lifestyle_context plsi state show --instance <instance> --map care_context plsi state show --instance <instance> --map personal_context plsi state show --instance <instance> --map cardiometabolic ``` 2. **Fill gaps in ONE `question`** (then wait), only for what neither the state nor the request's `Known:` block holds: what and how much per day or week, for how long, earlier quit attempts and what happened, triggers, and for alcohol: drinking on waking, shakes, sweating or confusion when not drinking, earlier withdrawal seizures. Answers go in your `result` as `lifestyle` items for `processor/intake`, and the amounts also as measured items on `lifestyle_context` (`alcohol_drinks_week`, `alcohol_days_week`, `cigarettes_day`; for alcohol, the AUDIT-C answers as `audit_c` when the user gave all three). 3. **Unsupervised stop is unsafe** with heavy daily drinking or any withdrawal sign: the plan's first action is a clinician visit (`test`) before cutting down; a past withdrawal seizure or confusion adds a `red_flag:` line (`orchestrator/urgent`). 4. **Evidence at run time**, verified, for the schedule and supports (counselling, quitlines, apps, `<cessation aid>`): ```bash plsi evidence search --source pubmed --query "<smoking cessation|alcohol reduction> <support> randomized" --limit 10 plsi evidence verify --kind pmid --id <pmid> ``` Reuse the evidence library and verify only what you cite (`shared/team`, Evidence speed). 5. **Dose** (`category: other`): a quit or reduce date within two weeks, a stepwise schedule (amount per day or week, alcohol-free days), plans for the top three triggers, and support by type only. A drug option goes in your proposal as `<cessation aid>` and is screened in `nostradamus/plan`; never name it here. 6. **Predict** on the amount itself: `alcohol_drinks_week`, `alcohol_days_week` or `cigarettes_day` on `lifestyle_context` at the schedule's steps and its end, from the recorded baseline, with a no-action control. No recorded amount yet: the first action is to record it (step 2). Predict on other coordinates only where the cited evidence measured them over the horizon (e.g. `sbp`, `resting_heart_rate`), as `change_from_baseline`. 7. **Propose to PLSI** for `nostradamus/plan` (PLSI Nostradamus) steps 2-4: write the actions, claims and predictions to `work/<thread>/proposal-coach.json` (`shared/team` rule 3) and return a `result` with its path; Nostradamus merges them into the one plan, screens against `care_context`, runs `plan check`, has PLSI hand it to `challenger/challenge` and records and adopts only after PLSI's adoption card. Never check, record or adopt a plan yourself. ## Rules - No product or brand names; support by type only. - A slip is recorded as adherence, never as failure; the review proposes `continue` or `modify`. - Never judge; numbers and the user's own reasons only.
© Insilico-org · MIT · Insilico-org/plsi