ritual
Scans your shell history, repos, and Claude Code memory. Ranks your top 5 automation candidates. Drafts the scheduled trigger that fixes it.
You're wasting hours on work that should be scheduled. The problem isn't motivation — it's visibility. You can't automate what you can't see. ritual scans your machine for repeated patterns and ranks the highest-leverage automations.
The Problem
Every founder runs the same five commands every Monday morning. Every operator pastes the same prompt into Claude six times a week. The patterns are obvious in retrospect — invisible while you're in them.
What It Does
- 01Audits your shell history, recent repos, and Claude Code memory in one pass
- 02Ranks your top 5 automation candidates by hours-saved-per-week
- 03Drafts a paste-ready Claude Code scheduled trigger for #1
The problem ritual solves
You know you should automate more. You don't because you can't see the patterns. Your shell history scrolls past, your repo work blurs together, and the same five tasks happen every week without anyone noticing.
ritual makes them visible.
What it does, end to end
Run it once and ritual will:
- Read your shell history for the last 30 days. Surfaces repeated command sequences.
- Scan your recent git repos for similar work patterns. If you ran the same
npm run+ commit + push sequence 12 times in a month, ritual catches it. - Read your Claude Code memory for prompts you've reused. Same prompt, different brand? That's a scheduled trigger waiting to happen.
- Cross-reference the patterns. Rank by estimated hours-per-week saved if automated.
- Draft trigger #1 as paste-ready Claude Code scheduled trigger syntax.
The output is a ranked list of automation candidates. Number one comes pre-drafted.
Example output
RITUAL — Top 5 Automation Candidates
1. ⭐ Run voice-scorer against /blog/ for every brand every Friday
Estimated time saved: 2.5 hrs/week
Draft trigger ready in `~/ritual/draft-1.md`
2. Render new social cuts from latest blog post (Mondays)
Estimated time saved: 1.8 hrs/week
3. Audit new client URLs with whystrohm-audit (on demand)
Estimated time saved: 1.5 hrs/week
4. Update CLAUDE.md voice profile from latest brand assets (monthly)
Estimated time saved: 1.2 hrs/week
5. Cross-post Slop Files episodes (on episode publish)
Estimated time saved: 1.0 hrs/week
Total potential: ~8 hrs/week
Why scheduled triggers specifically
Anthropic shipped scheduled triggers for Claude Code in early 2026. They run remotely, on Anthropic's infrastructure, on a cron schedule you define. You don't have to leave your laptop open.
ritual is built specifically to find work that fits this pattern: recurring, well-defined, machine-completable. It doesn't try to automate creative work or one-off problem-solving. It targets the boring, repeated, definitely-shouldn't-be-manual work.
Install
Download the latest .skill from GitHub releases. Drop it in ~/.claude/skills/. Restart Claude Code.
# Find the latest release
open https://github.com/whystrohm/ritual/releases/latest
What you get on first run
A markdown report at ~/ritual/audit-{date}.md with the ranked list, plus a draft-1.md containing the paste-ready trigger for your #1 automation candidate.
Edit the draft if you want, then paste it into Claude Code as a scheduled trigger. Done.
Full docs and methodology on GitHub →
How It Composes
ritual is the orchestration layer that sits above every other skill. Once it identifies a routine worth automating, it can schedule any of the other WhyStrohm skills as a scheduled trigger — voice-scorer running nightly, audit running on every new blog post, voice-extract running weekly against competitors.
Related Skills
media-tsunami
The empirical layer. Extracts brand voice as executable code — cadence, vocabulary, forbidden words, exemplar sentences — serialized as a CLAUDE.md any LLM can load.
Install →whystrohm-audit
Score any content against a 5-layer infrastructure framework. Diagnose what's broken. Rewrite one piece live.
Install →