feat(skills): lawb-pool-monitor — autonomous prize-pool watcher for LawbWorld arcade on Base

[lawbworld-tech]

Summary

New AEON skill watching the LawbFishing prize pool on Base mainnet — the on-chain economy behind the LawbWorld LAWBFISH arcade (https://lawb.world). Hourly read-only run, threshold-gated alerts, silent unless
operator attention is needed.

Replaces a manual workflow that previously required SSHing into a dev machine and running npx hardhat run scripts/diag-fishing.js repeatedly — sometimes 10+ times per day during heavy redeem activity.

What it watches

• prizePool() — LAWB available for redemption
• shopVault() — should stay 0 post-upgrade; alerts if accounting drifts
• paused() — emergency stop
• 24h rolling burn-rate window
• Recent Redeemed events for activity context

Alert conditions (6h cooldown per condition)

Condition	Trigger
paused	contract paused (critical)
shop_vault_nonzero	accounting drift detected
low_pool	< 10M LAWB
critical_low_pool	< 1M LAWB
high_burn	> 100M LAWB/24h

Real-run output (Telegram)

[screenshot you just shared]

Agent improvised an extended 5h-window scan when the 1h sample had no activity, and added contextual framing for the shop_vault_nonzero alert ("today is the documented upgrade date — worth a manual check"). Quality
way above the bare spec.

Files

• skills/lawb-pool-monitor/SKILL.md — the skill spec (read-only, no on-chain writes)
• aeon.yml — registers skill under crypto/midday section, disabled by default

Context

LawbWorld is a Base-native NFT + game ecosystem. LawbFishing proxy at 0x48b2db9E89542Baa217bf8dc6269164b7887fE57 (verified). The skill is contract-version-stable — selectors are baked in and recomputable via
keccak256(<sig>)[:4] if redeployed.

Happy to iterate on the format, the alert thresholds, or the notification body shape based on what fits the AEON skill style best.

[imancipate]

Triage: OUT-OF-SCOPE — touches scripts/postprocess-devto.sh (mode change 100644→100755), a protected scripts/postprocess-* path external contributors cannot modify directly. The PR also commits agent-runtime artifacts (.outputs/, dashboard/outputs/, memory/) that shouldn't ship in source.

The new skill itself (skills/lawb-pool-monitor/SKILL.md + aeon.yml registration) is the substantive part — if you drop the chmod and the runtime artifacts and keep only those two files, this re-triages as ACCEPTED on the next push (rubric otherwise clean: format ✓ / originality ✓ / size 464≤500 ✓). Leaving open and labelled triage:out-of-scope; @aaronjmars can override if the scope read here is too strict.

[lawbworld-tech]

Hi @aaronjmars — apologies for the noisy PR. The substantive change is just aeon.yml (+1 line skill registration) and skills/lawb-pool-monitor/SKILL.md (227 lines). The other 6 files are runtime artifacts (.outputs/,
▎ dashboard/outputs/, memory/) and an accidental mode change on scripts/postprocess-devto.sh — none of those should ship.
▎
▎ I'm temporarily locked out of push access on this fork (email account suspended, can't re-auth gh/generate PAT). Could you either override the triage:out-of-scope label, or close this PR so I can re-submit cleanly once access is
▎ restored? Happy with either.