Files
winning-wallet-finder_github/LIVE_TEST.md
T
jaxperro 8c07343735 discord: retire per-trade pings, add daily sharp digest; trust: v2 self-certify + invalidated-wallet fallback
* per-trade Discord pings removed everywhere: copybot placement/settle/startup
  alerts silenced (engine.webhook forced off), webhook_receiver.py archived,
  Procfile removed. The ONLY Discord output is now live/discord_daily.py at the
  end of the daily pipeline: one embed listing every watch_sharps wallet with a
  Polymarket profile link + 30D conviction win% / record / P&L.

* trust.py fixes found on the first scheduled run (sharp list collapsed 25->7):
  - Polymarket rewrites endDate after resolution, so freshly re-pulled wallets'
    corrected res_t stopped matching the stale cross-wallet consensus
    (374/454 of iohihoo's rows wrongly distrusted). v2 rows with resolved=TRUE
    are now self-certifying (resolved=TRUE implies endDate-based res_t observed
    post-end) — consensus only gates legacy rows.
  - daily.sh invalidates watchlist wallets by DELETING their pulled row; a
    transiently failed re-pull then dropped the wallet's entire history from
    the trusted set (the 0x73afc816 whale vanished). Legacy rows of wallets
    missing from pulled now fall back to trusting markets resolved >=14d ago.
  Re-run: 85 conviction wallets, 34 copy-positive sharps — whale/iohihoo/
  Stavenson/LSB1/Kruto all recovered.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-04 18:31:30 -04:00

3.0 KiB
Raw Blame History

Live pipeline test — minimum-size real-money runbook

Goal: a few $5 trades through the full pipe (detect → filter → size → place → settle → redeem) with hard caps bounding worst-case loss to the deposit.

Caps in config.live.json (gitignored — holds your key, never commit): $5/trade · $25/day · $30 max exposure · 6 open positions · 2 per event · prices 0.050.95 · FAK marketable orders. Worst case ≈ the $30 exposure cap.

1. Fund a dedicated account (you)

  • Use (or create) a Polymarket account you're comfortable putting a key on this machine for — deposit only the test money (~$50). The key controls only what's in the account; keep it that way.
  • Deposit USDC to the account (the profile/deposit address = your funder).
  • Send ~1 POL to the signing EOA for redeem gas (skip if you set live.auto_redeem: false and redeem manually in the UI).
  • Export the private key: email-login accounts → Polymarket Settings → Export private key (signature_type: 1, already set). Browser-wallet accounts → your wallet's key, and set signature_type: 2.

2. Fill config.live.json (you)

"live": {
  "private_key":    "0x…",   // the exported key
  "funder_address": "0x…",   // your Polymarket profile/deposit address
  "signature_type": 1        // 1 = email login · 2 = browser wallet
}

3. Preflight (read-only, no orders)

python3 preflight_live.py

Green across the board = auth, USDC balance, market access, and redeem gas all verified. Fix anything red before arming.

4. Pause the paper worker (avoids the two bots fighting over the feed)

cd ~/…/scratchpad/railway-copybot && railway down --service copybot -y
# …or pause the service in the Railway dashboard

5. Arm (you type the phrase — that's the point of it)

python3 copybot.py --config config.live.json \
    --state copybot_state.live.json --poll 60 --live

It prints the caps, asks for the confirmation phrase, then baselines (no retro-copying) and waits for the next conviction trade from Kruto2027 / shisan888 / fortuneking / LSB1. Placements log to the console/Railway logs (per-trade Discord pings were retired 2026-07-04). Watch the first fill: order accepted → shares in the account (Polymarket UI) → on resolution, SETTLE line + auto-redeem tx hash.

6. Wind down

Ctrl-C the live bot, then redeploy the paper worker:

cd ~/…/scratchpad/railway-copybot && railway up --service copybot --detach

Notes

  • Venue: the bot trades the international CLOB (clob.polymarket.com). If your account lives on the US exchange, its API differs — confirm which venue your funds are on before the test.
  • Neg-risk markets can't auto-redeem — the bot warns; redeem those in the UI.
  • The live book/state is separate (copybot_state.live.json); the paper July test's state is untouched.
  • First redeem.py run is untested against live — verify the first redemption tx on Polygonscan before trusting the loop.