bbrysonelite-max Brent's Daily Checks
Portfolio-level morning ritual: every business gets its own evidence block, and Brent gets one top-line verdict. Never blend evidence across businesses — a Tiger failure is not an Alien Probe failure, and an Alien Probe HOLD does not make Tiger unhealthy.
Boundaries
- Report and flag. Do not auto-fix unless Brent separately says
fix it. - One child failing never blocks an independent child. Mark unavailable children
UNVERIFIEDand name the exact missing dependency. - Never blend Tiger customers, leads, Mine data, or incidents into Alien Probe product evidence.
Receipts
Every live claim = command + output + UTC stamp, or the row reads UNVERIFIED. No relative time words ("earlier", "just now") as proof.
0. Integrity — always first
Read the actual bytes of:
~/.claude/skills/brents-daily-checks/SKILL.md~/.claude/skills/alienprobe-product-puck/SKILL.md~/.claude/skills/tigerclaw-daily-checks/SKILL.md~/tiger-claw-v4-core/DAILY_CHECKS.md
Verify the scope split:
- This parent references both child skills.
- Tiger's skill and canonical procedure contain no Alien Probe planning step.
- Alien Probe's skill contains no Tiger operational procedure.
- Tiger's
daily:checkspackage entrypoint exists (~/tiger-claw-v4-core/api/package.json). - All three frontmatter blocks contain only
nameanddescription.
Do NOT run the Python skill-creation validator as a daily dependency — it is a development-time check with its own managed environment. Daily integrity must never fail merely because the system Python lacks PyYAML.
If the split is missing or contradictory, put BRENT DAILY CHECKS DISARRAY first. Continue any child that is present and readable; never return an all-good verdict.
1. Alien Probe — product puck
Run /alienprobe-product-puck in Daily mode. Preserve its five lines exactly:
alien_probe_plan HOLD | ADVANCE | INVESTIGATE
shelf <live products, prices, and probe result>
buyer_proof <independent buyers, repeat buyers, verified revenue, or UNVERIFIED>
new_signal <one meaningful delta or "none">
planning_action <one bounded research/planning action; never a build action>
Planning and read-only measurement only. Do not pay, settle, build, deploy, list, contact a buyer, or touch credentials.
2. Tiger Claw — operational health
Run /tigerclaw-daily-checks end-to-end against its canonical procedure. Keep its result under its own Tiger Claw heading.
Tiger /health is not a substitute for "FirstTouch actually sent." Green health with no sent= in the last 24h is a flag.
3. Rest of business (Brent 2026-08-19)
Separate heading, after the two children. Do not blend these numbers into the Tiger or Alien Probe blocks. Write the filled page to ~/Desktop/HANDOFF/YYYY-MM-DD Daily Check/ (created per run).
If nobody was contacted yesterday → verdict NO MONEY even if Tiger is green.
| # | Check | Want |
|---|---|---|
| 1 | Money | $ in yesterday · conversations · next cash move |
| 2 | Tiger sends | FirstTouch / ReferralCadence sent= · USAGE_LIMIT / CAP_REACHED |
| 3 | Mine | facts/24h · newest fact · 0% pollution · unleased NM pool |
| 4 | Leads | new named · contacted · right kind · they answered · next call on calendar |
| 5 | CLAW | a stranger (not Brent) commented CLAW and got https://stan.store/brentbryson/p/5-ai-agents-that-will-3x-your-business |
| 6 | WTH | 3 posts fired on X |
| 7 | Horror Stories | NM scar campaign (internal endless-referrals) · IG/FB live or scheduled |
| 8 | YouTube longform | DEAD DEAL. Report views only. Do not produce, thumb, or publish. |
| 9 | YouTube Shorts | public scar Shorts vs unlisted glossary |
| 10 | Rebuild | which of 6: CSV export → Tiger import → backup → point at builders → run → log wins |
| 11 | Consulting | Pat thread · billed work |
Right-kind lead: NM builder/leader, struggling company or looking for a vehicle. Not Debbie. Not a stadium. Teaching YouTube: Brent's voice. Second voice only if a few words — the dual-clone format failed the couch test.
Mark any row you could not ground UNVERIFIED.
Verdict
ALL GOOD— integrity passes, every child healthy, at least one human contacted yesterday, no grounded failures.NO MONEY— nobody was contacted yesterday, even if Tiger and Alien Probe look fine.ISSUES— one or more grounded failures.PARTIAL— one or more checks could not be grounded.DAILY CHECKS DISARRAY— the parent/child split or a required procedure contradicts itself.
Report in this order, most consequential issue first. Short lines, no walls, no raw logs or secrets unless Brent asks for a specific non-secret excerpt.
BRENT'S DAILY CHECKS — <verdict> — <UTC timestamp>
daily_checks_integrity <ok | DISARRAY + exact reason>
Alien Probe
<five-line product-puck result>
Tiger Claw
<Tiger daily-check result>
Rest of business
1 Money … 11 Consulting (one line each)
first_action <one bounded next action or "none">
Dependencies
Mandatory: the three installed skill files above; ~/tiger-claw-v4-core/DAILY_CHECKS.md; Tiger's daily:checks entrypoint; and this section-3 list (also on Desktop as Daily Check List.md). Each child owns its own live-service, auth, and repo dependencies.