Morning Boot

Morning Brief — [Day, Mon DD · HH:MM ET]

🟢 [System]: [one phrase] 🟡 [System]: [one phrase — what's attention-worthy] 🔴 [System]: [one phrase — what needs action] 🎯 Objectives: [GREEN|YELLOW|RED] — unmapped=[N], built_unmapped=[N], high_pri_unmapped=[N]

Top 3 to Grow This Week

1. [Opportunity] — [why it compounds, what the lever is]
2. [Opportunity] — [why it compounds, what the lever is]
3. [Opportunity] — [why it compounds, what the lever is]

→ Verbose details: Kai Morning Briefing.md Blog jam (append if outline is ready) If blog_outline in the briefing has status ready , append ONE line after the Top 3: 📝 Blog jam ready: "[theme in 8 words or less]" ([signal]) — say "let's jam" to draft it. Do NOT show this line if no outline exists. Keep it short — it's an invitation, not a report. Traffic light rules Color Meaning Nick's action 🟢 Working as expected — no attention needed None 🟡 Worth watching or recently changed — awareness only Mental note 🔴 Blocking revenue, broken, or actively losing ground Decide today Objectives line mapping: 🎯 Objectives: RED if built_unmapped > 0 OR high_pri_unmapped > 0 🎯 Objectives: YELLOW if only unmapped_total > 0 🎯 Objectives: GREEN if all objective counts are clean Systems to always cover (in this order): Revenue gates (🔴 if any open) Live site / auth X reach / content Compass (build ratio, momentum) Objectives alignment (mapping hygiene + built/high-priority unmapped) Team / EOD Pipeline health Memory layer (🧠) — GREEN if corpus fresh <8h, YELLOW if <24h, stopped if service down (stopped = 🟡 only, not 🔴) What makes a 🟡 vs 🔴: 🔴 = money is leaking, users are broken, or we're losing ground TODAY 🟡 = recently fixed (watching), slightly off-target, or needs a decision this week If in doubt: 🟡. Reserve 🔴 for things that cost money or users right now. Top 3 rules These are not the 3 biggest problems. They are the 3 highest-leverage moves. Each Top 3 item must be efficiency-positive: Faster cycle OR same effort with materially higher throughput/quality Lower cost per outcome Lower coordination burden Ask: "If we did this one thing consistently this week, what would compound the most?" Framing: ✅ "Auto-schedule content — queue is ready daily but nothing ships; fixing this = consistent reach with zero extra effort" ✅ "Trial→paid conversion — 46 trials/day at $0 conversion; one email sequence = direct revenue" ❌ "Fix the Resend API key" — that's maintenance, not growth (put it in 🔴 instead) One sentence each. What it is, why it compounds. No bullet nesting. Rotate the lens each day: Mon = content/reach, Tue = conversion/revenue, Wed = product, Thu = team leverage, Fri = ops/systems. Don't always surface the same 3. Phase 6.5: Engineering Plan Card After the surface report, surface the engineering plan + Jules budget. Read it from: step_results["engineering_plan"] (already in briefing) — do NOT re-run it. Budget from: python3 /Users/nicholaspetros/scceo-1/grok-personal/jules_budget.py Format:

🔧 Engineering Plan — Kai Drives Today

{ranked list from engineering_plan["plan_md"]}

📊 Jules budget: {used}/{300} used today ({remaining} remaining) 30-min pulse active — status posts to team chat as items complete.

Say "dispatch plan" to generate task cards + send all {N} items to Jules. When Nick says "dispatch plan": cd /Users/nicholaspetros/scceo-1/grok-personal && python3 jules_batch_dispatch.py This will: Generate a task card for each item (from work ledger metadata) Write task card to {repo}/plans/{item-id}.md + scceo-1/plans/{item-id}.md Dispatch to Jules ( jules remote new --repo {github-repo} --session ... ) Update work ledger: status → dispatched, record session_id Debit Jules budget Budget rules: Hard stop at 280/300 — reserves 20 for ad-hoc/Nick dispatches Soft warning at 250 — surface in pulse but keep dispatching 300/day resets at midnight UTC After dispatch: Heartbeat pulse will monitor every 30min and post to team chat when: A Jules session completes (PR URL available) Budget hits warning/hard stop Escalations need attention Rules: Only show this block if engineering_plan["status"] == "success" and there are items If engineering_plan is missing or failed: skip this block silently Do NOT add commentary or re-rank items — show them exactly as generated The goal: Nick reads it in 10 seconds, says "dispatch plan", then focuses on growth Nick saying "dispatch plan" IS the explicit override for plan approval (no Grok review cycle needed) Phase 6: Hand Off After the surface report: If 🔴 items exist → "Want me to queue a fix for X?" (don't fix automatically unless obvious) If top 3 resonate → "Want me to dispatch on any of these?" If Nick says "let it rip", "dispatch", or "start the conveyor" → run the conveyor-dispatch skill (full morning dispatch: sync → select → generate task files → enqueue to heartbeat) If Nick says "let's jam", "jam on the blog", or "blog draft" → read the blog-outline.md file and run the morning-content-jam skill Phase 3 jam flow: prompt Nick section by section using the speak-to prompts, collect his answers, then draft the article inline using his voice profile. After the draft is approved/published, run: python3 /Users/nicholaspetros/scceo-1/grok-personal/blog_published.py to reset the streak nag. Always end with: "What do you want to start on? Say let it rip to run the full conveyor dispatch." Phase 7: Skills Menu After the hand-off line, always append a compact skills block. Two parts: suggested picks for today + a standing reference card . Suggested picks (rotate by day) Always anchor to morning-content-jam if queue is ready. Add 1-2 others by day: Day Suggest Mon morning-content-jam · sos-pulse · repo-planning Tue morning-content-jam · codedoc-scan-check · cto-heartbeat Wed morning-content-jam · run-testing-suite · status-reconcile Thu morning-content-jam · sos-pulse · daily-plan Fri morning-content-jam · codedoc-scan-check · rocket-ci-gates Sat/Sun morning-content-jam · daily-plan · grok-plan-loop Context overrides (add these regardless of day): Heartbeat queue has pending items → suggest cto-heartbeat Work ledger has failed/unverified gates → suggest status-reconcile Blog outline ready → say "say 'let's jam'" not a skill name No test run noted recently → add run-testing-suite Output format Append this block at the very end, after the hand-off line:

Skills to run this morning: → morning-content-jam — punch up today's reply queue → codedoc-scan-check — dead code + doc hygiene → cto-heartbeat — dispatch queue + completions

All skills: Content morning-content-jam Dispatch conveyor-dispatch · cto-heartbeat · plan-advance Planning daily-plan · repo-planning · grok-plan-loop Health sos-pulse · codedoc-scan-check · run-testing-suite · status-snapshot · status-reconcile CI/Deploy rocket-ci-gates · app-ci-gates · scceo-ci-gates The reference card stays the same every day. Only the "Skills to run this morning" picks change. Rules Obsidian Kai Morning Briefing.md gets everything. The session gets the summary. Never repeat the full briefing in chat. Top 3 are growth moves, not the 3 worst problems. Green things compound; broken things get queued. Revenue gates are check #1. If a gate is open it goes in 🔴 and leads the report. Run objective audit daily from Morning Boot. Do not rely on Kai Morning usage. If built/high-priority items are objective-unmapped, surface as 🔴. Don't wait for the pipeline to finish. Show health lights first (fast), pipeline output second (slow). $30K plan trajectory belongs in the compass line. Every morning: are we on pace? "Needs Nick's attention" ≠ "broken." 🟡 is for things worth knowing, not just things that failed. Zero emails in 24h = 🔴. Always. Even if expected — surface it so Nick decides. Keep it to one screen. If it doesn't fit without scrolling, it's too long. Python Automation Audit After boot: were any morning check steps deterministic (gate query, health ping, file parse, metric pull)? Quick check: python3 ~/.agents/skills/prefer-python-over-llm/scripts/classify_task.py "step description" If scriptable → log rule_ledger.skill_usage with python_candidate=true and flag for cron replacement Rule: ops-prefer-python-over-llm Variants / Absorbed Modes --check Daily status check only — gates, nag items, pulse. Skip content generation. (absorbed: daily-check) --plan Generate today's day plan from priorities and work ledger. (absorbed: daily-plan) --status Quick current status across all systems. (absorbed: get-status) --next Show what to work on next based on unblocked priority items. (absorbed: whats-next) --dashboard Check dashboard health indicators only. (absorbed: check-dashboard-status)