skills/kairos-shutdown/SKILL.md
End-of-day capture — ClickUp completion check, uncommitted work, intensity, tomorrow's frog. Use when closing out the workday.
npx skillsauth add mpazaryna/agentic-factory kairos-shutdownInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Quick end-of-day ritual - under 3 minutes. Capture and clear.
If the target day is Saturday or Sunday, skip. No weekend shutdowns.
Before gathering context, determine which day this shutdown is for.
## Shutdown section.## Shutdown section.If the previous working day's note exists but has no Shutdown section: This is likely a late shutdown. Confirm with the user:
Looks like [previous day] didn't get a shutdown. Closing out that day or today?
Wait for response. The answer determines the target day.
If both have Shutdown sections or the previous day has one: Target day is today. Proceed normally.
If the user explicitly says "shutdown for [date]": Use that as the target day.
Track whether this is a late shutdown (target day < today). This affects the frog question later.
Read in parallel, don't output yet:
kairos/logs/weekly/) - what was planned for today? What's planned for tomorrow? Are encouraged projects intentionally skipped this week?projects/*.md - all non-archived projects~/workspace/primary-pm/_tools/clickup-today.sh [DATE] --shutdown to get both open AND completed tasks (including subtasks closed today without a start_date)Run git status in repos for active projects. Only surface if there are uncommitted changes.
If found: "resin-platform has uncommitted changes - commit, stash, or leave?"
Don't use AskUserQuestion. Just ask and wait for response.
Run ~/workspace/primary-pm/_tools/clickup-today.sh [DATE] --shutdown to get the full picture:
Surface the summary: how many completed, how many still open. The ClickUp data tells you which projects got attention -- don't ask the user to repeat this.
ClickUp: 11 of 13 tasks completed today. Open: "Pipeline Phase 2", "Sean Allen Course" -- carry to tomorrow or reschedule?
Simple, conversational. No predefined options.
ClickUp data already tells you what was accomplished and which projects got attention. Don't ask the user to repeat what the data shows. Instead, present the summary and ask only what the data can't tell you:
ClickUp shows [N] tasks completed today across [projects].
[list key completions]
[N] still open: [list] -- carry to tomorrow or reschedule?
Any blockers or carry-over beyond those?
Any project status changes? (unblocked, stalled, ready to archive?)
Quick metrics for the load system:
- Intensity today? (1-5, where 1=coasting, 3=steady, 5=full burn)
- How many blocks did you actually work?
These metrics feed the load calculation that determines next week's pace.
After capturing today's intensity, check the previous working day's frontmatter for its intensity value.
If today is intensity 5 AND the previous day was also intensity 5:
Two burn days in a row. You've seen this pattern before --
it leads to depletion, not throughput. Tomorrow must be a 3
or lower. What can you cut or simplify from tomorrow's plan?
Don't accept "I'll be fine" or "I'll see how I feel." The data is clear: consecutive intensity-5 days produce a crash. A deliberate recovery day is not lost productivity -- it's what prevents losing the next three days. Marathon, not sprint.
If today is intensity 5 (but the previous day wasn't):
Note it without alarm, but seed the awareness:
Intensity 5 today. Keep tomorrow at 3-4 to avoid the
consecutive burn pattern. What's a lighter plan?
Wait for responses, then record.
Update the target day's daily note -- Shutdown section AND frontmatter:
Frontmatter additions:
---
tags: [daily]
date: 2026-03-03
intensity: 4
blocks: 3
projects: [chiro, resin]
---
Shutdown section:
## Shutdown
**Accomplished:**
- [from response]
**Carry-over:**
- [from response, or "None"]
**Project Changes:**
- [any status changes, or omit if none]
**Tomorrow:**
- [from response]
**Load:** intensity 4 · 3 blocks · chiro, resin
Late shutdown note: If this is a late shutdown, add annotation:
## Shutdown (recorded Mar 4)
Before closing, ask about the frog. Adapt based on timing:
Same-day shutdown:
Name tomorrow's frog. One specific task -- not "admin."
(Or "nothing pending")
Late shutdown (running the next morning):
Name today's frog. One specific task -- not "admin."
(Or "nothing pending")
Friday shutdown: Tomorrow is the weekend. No frog needed. But ask:
Anything lingering that should be Monday's frog?
If they name something, include it in the Tomorrow section with the label Frog:. This seeds the next kickoff.
**Tomorrow:**
- **Frog:** Follow up with Joe on billing
- Chiro UAT continues
Brief summary of what was captured:
Same-day shutdown:
Recorded. Uncommitted work handled. See you tomorrow.
Late shutdown:
[Day] closed out. Recorded to [date]'s note. Ready for today.
Friday shutdown:
Week captured. Run /kairos-review weekly for full retrospective. Have a good weekend.
development
Comprehensive Cloudflare platform skill covering Workers, Pages, storage (KV, D1, R2), AI (Workers AI, Vectorize, Agents SDK), feature flags (Flagship), networking (Tunnel, Spectrum), security (WAF, DDoS), and infrastructure-as-code (Terraform, Pulumi). Use for any Cloudflare development task. Biases towards retrieval from Cloudflare docs over pre-trained knowledge.
tools
Send and receive transactional emails with Cloudflare Email Service (Email Sending + Email Routing). Use when building email sending (Workers binding or REST API), email routing, Agents SDK email handling, or integrating email into any app — Workers, Node.js, Python, Go, etc. Also use for email deliverability, SPF/DKIM/DMARC, wrangler email setup, MCP email tools, or when a coding agent needs to send emails. Even for simple requests like "add email to my Worker" — this skill has critical config details.
tools
Build AI agents on Cloudflare Workers using the Agents SDK. Load when creating stateful agents, durable workflows, real-time WebSocket apps, scheduled tasks, MCP servers, chat applications, voice agents, or browser automation. Covers Agent class, state management, callable RPC, Workflows, durable execution, queues, retries, observability, and React hooks. Biases towards retrieval from Cloudflare docs over pre-trained knowledge.
tools
Planning conductor — runs the full PRD → Spec → Gherkin loop in a single interactive session with human approval gates.