Detailed content
The Cadence test
There is a simple, unforgiving test to see whether Cadence is working: close your laptop. If a brief arrives in your inbox and a real response emerges without you typing anything — your AIOS has passed. Cadence isn't about automating; it's about presence without presence.
🔑 Main Concept
Cadence is proof that the three previous Cs work. A system that only acts when you’re present isn’t autonomous — it’s an expensive tool to call.
The closed laptop is the acceptance criterion. When a colleague asks a question and gets a real answer from your AIOS—without you stepping in—the system has proven its operational value.
✓ What to DO
- ✓Test the system with the laptop closed before declaring that it's working
- ✓Let a real colleague’s question reach AIOS and observe the response
- ✓Calibrate the autonomy threshold (what it can answer on its own vs. escalate)
✗ What NOT to do
- ✗Consider “Cadence active” just because you configured a hook — testing is required
- ✗Give generic answers: AIOS needs to know the real context of the business
- ✗Automating without first validating that the workflow works with you present
Recurring triggers
The AIS-OS kit uses two mechanisms to trigger behavior without manual interaction: hooks in settings.json e skills named by temporal pattern. These two mechanisms form the infrastructure for Cadence—but remember: infrastructure doesn’t replace ritual.
💾 Naming convention — Temporal Skills
daily-* → roda todo dia (ex: daily-standup, daily-digest)
weekly-* → roda toda semana (ex: weekly-review, weekly-kpis)
monthly-* → roda todo mês (ex: monthly-retro, monthly-goals)
morning-* → rotina matinal (ex: morning-brief, morning-check)
standup → síntese para sync de equipe⚙️ Hook on .claude/settings.json
{
"hooks": {
"PostToolUse": [{
"matcher": "Write",
"hooks": [{
"type": "command",
"command": "node scripts/daily-digest.js"
}]
}]
}
}PostToolUse fires after every tool use. Use sparingly — too many hooks create noise.
💡 Practical Tip
Start with morning-brief — a skill that summarizes what’s on the agenda today. It’s the most natural, lowest-risk trigger for validating that Cadence works before more complex automations.
Recurring rituals
The AIS-OS kit’s anchor isn’t cron. It’s the ritual: consistent behavior you perform until it becomes a reflex. Two rituals form the operator's weekly cycle—and together they cover diagnosis + growth.
/audit — Weekly diagnosis
Every seven days — the day of the week doesn’t matter; the frequency does
Run the Four-Cs in your current installation. Ask: Does Context still reflect the business? Are Connections broken? Are Capabilities outdated? The output of the /audit is the input for the next ritual.
/level-up — Weekly progress
Every Friday, 15 minutes — before you shut down your computer
An automation, a new skill, a connection added. The /level-up doesn’t need to be grand — it needs to be consistent. 1% better per week = 2× better in 70 weeks.
⚡ Behavior > Infrastructure
An AIOS with inconsistent rituals but good infrastructure is less effective than an AIOS with simple infrastructure and rituals performed every week. Habit is the hardest component to replicate — and the most valuable.
Cadence comes LAST
There is a dependency graph across the 4 Cs — and it is rigid. Context is the non-skippable foundation. Connections and Capabilities can be built in parallel. But Cadence only holds up when the three previous ones are in operation.
⚠️ Cadence’s Golden Rule
"Don’t automate workflows that don’t work manually."
Automation amplifies. If the manual workflow is fragile, automation delivers fragility at scale. If it’s solid, it delivers solidity at scale. The order matters: validate manually first, automate afterward.
✓ Correct sequence
- 1.Context: document the business (CLAUDE.md, .context/)
- 2.Connections: connect tools and data sources (MCP, integrations)
- 3.Capabilities: configure skills and workflows that work manually
- 4.Cadence: now automate the validated rituals
✗ Shortcuts that break things
- ✗Create a cron job for a task that has never been done manually
- ✗Automating a daily digest without documented Context — useless output
- ✗Skip Context and go straight to Cadence — autonomous system without memory
The 3 indicators of success
The 3 indicators of success aren’t KPIs—there’s no dashboard and no numerical target. They are lived experiences that signal AIOS has irreversibly changed how you work.
Team-reaches-out
Indicator 1 of 3
Colleagues start asking you things about the system — "how did you do that?", "can I use your template?" Or better: the system answers for you, and the person doesn't even realize you weren't there. The team starts orbiting the AIOS, not you.
Context-switching reduction
Indicator 2 of 3
You stop being the context bottleneck. You no longer need to explain the same project to five people. AIOS carries the context and distributes it. You find yourself spending less time in "explanation" mode and more time in "decision" mode.
Knowledge-leaves-your-head
Indicator 3 of 3
The knowledge that used to exist only in your head now lives in the system. When you take time off, a project doesn’t stall. When a colleague needs help, the AIOS provides it. Institutional knowledge has left the silo and become part of the infrastructure.
💡 How to monitor without a dashboard
Every Friday in the /level-up, ask: "Has anyone come to me for something the AIOS should answer?" and "Did any knowledge leave my head this week?" If the answer is no for three weeks in a row, go back to Cadence and review the rituals.
Personal → AI-ready company
When all three indicators appear for one person, something important happens: the same architecture scales. The data, the connections, the capabilities — everything you’ve built personally becomes the blueprint for the team and the company.
🚀 The scale effect
A company where every operator runs a personal AIOS is truly an AI-ready company. Not because it adopted an AI platform, but because each person knows how to operate autonomously, and institutional knowledge lives in infrastructure, not in people’s heads.
Scale happens in three layers: dashboards (visibility), automations (execution) and team rollout (replication). Each layer assumes the previous one is working.
Dashboards
Visibility into what AIOS is doing. Business KPIs connected to the system. Lets you observe before scaling.
Automations
Workflows that once depended on human intervention now run through hooks, recurring skills, and established integrations.
Team rollout
The personal AIOS becomes a template for other operators. Each onboarded person starts with an architecture already validated by the pilot.
⏰ Module Summary
Next: Track 3 — Operate in Practice
You’ve completed the 4 Cs of architecture. Now it’s time to put everything into operation: /onboard, /audit, /level-up, and real expansions in your business.