Monday edition. Reviews Mon Oct 5 → Sun Oct 11 (7 UTC days) against Sep 28 – Oct 4. Solana volume $4.97M vs $17.98M (−72%); $4.20M of it from ordinary wallets (−67%). The week had one event: Finn's Wed Oct 7 "anyone know what this is?" screenshot (571❤️ / 72K views in 8h). It produced a $2.03M day, 5,223 traders and 98 launching creators — the best organic-creator day in a month — and no follow-up. Thu $883K, Fri $301K, back where the week started. Platform take $21.8K / 189 SOL (−73%), DefiLlama revenue $22.8K (−73%), creator claims $36.0K to 121 creators (−75%). Launches 2,208 (−8%), but Sat 19 and Sun 35 are the lowest days on record, because 77% of launches are our own brrr.trade deployers and the launches floor has been off since Oct 6, the stock lane since Oct 8. The 797 stock pairs we launched Mon–Thu traded $147K. Eng shipped 214 human commits (+22%), 142 of them alaa's Postgres cutover (shadow sink, reconciler, datapi family reads, ~66K lines merged); #549 claim status merged Monday; the bot fleet (Dividend, Donation) went to production GitOps. Robinhood: 7.07 ETH ≈ $17.7K for the week (−80%), 2 launches, vault 0.0 ETH. X data is five days old: the API credits ran out on Oct 8–12 and every route 402s.
Operating frame: Bags runs as a KPI-focused firm aimed at capturing trading volume through mobile and terminal products. The stats come prefilled (this brief is the prefill); the meeting discusses why. This week's why: attention is the input and we rationed it to one screenshot. The receipt that was owed on Oct 8 still hasn't gone out, the holders' replies filled with the hackathon grievance, and on-chain the decay was textbook: $2.03M → $883K → $301K.
Key stats: Solana $4.97M (−72%), 12.6K traders (−59%), 83K trades (−88%), 2,208 launches (−8%) from 231 creators (−14%). Creator claims $36.0K / 121 claimers (−75%). Platform take $21.8K / 189 SOL (−73%) = public revenue $22.8K on DefiLlama (−73%). Take rate 0.44%, flat, still reconciled day-by-day. Robinhood 7.07 ETH ≈ $17.7K (−80%), 2 launches; vault 0.0 ETH. SOL fell to $107.7 (−10%), which is ~a third of the USD drop on its own. Eng: 214 human commits (+22%), 46 PRs merged (30 by humans), Linear closed 9. Headline: a comedown week, with one 36-hour proof that a single post from Finn is worth $2M of volume.
Solana by day (excl. high-frequency wallets): $247K → $231K → $1.86M → $806K → $276K → $415K → $363K. Daily traders 1.7K → 1.6K → 5.2K → 3.5K → 1.7K → 2.4K → 1.9K. $BAGSPAY was $1.10M of buy-side volume (13K trades, 3,290 makers), a fifth of last week's. New this week: $PEPPER (launched Fri Oct 10) $70.6K / 734 makers and $GROKBIT (Sat) $50K — the weekend was organic coins, not ours. Coins launched this week did $762K (15% of volume) across 1,766 mints. Launches: 54 farm wallets (≥20/wk) made 1,699 of 2,208 (77%); the five biggest are brrr.trade deployers at 48–49 each. Organic: 177 creators (−18%) launched 509 coins (+4%). Wed Oct 7 had 98 creators, the most in four weeks — the tease pulled builders, not just traders. Stock pairs: 797 launched (196 last week) → $147K traded, 1,232 buyers, $184 per pair. Then the lane died with the fbmini outages on Oct 8 (0 on Sat).
Robinhood: 7.07 ETH ≈ $17.7K over Oct 5–11 vs 35.5 ETH / $96K for Sep 28–Oct 4 (−80%), 2 launches vs 23. By day: 1.29 → 0.99 → 1.19 → 1.04 → 0.84 → 0.17 → 1.55 ETH. Curve 2.46 / pool 4.62 ETH, 83 buys / 96 sells / 121 pool swaps. Fees (2%) ≈ 0.141 ETH ≈ $355, protocol half ≈ $177. All-time launches 4,757 (+2 on the cache refresh). Vault 0.0 ETH. Two weeks post-subsidy, RH is a rounding error: ~1 ETH a day of pool swaps on old coins and no new launches.
The claim surface is live end-to-end. #549 (token-page claim status, +10.6K, finnbags) merged Monday; ramyodev indexed the claim vault events (#173), routed claim transactions to the claim feed (#691 mobile, #556 web), re-enabled claim tweets for every creator claim type (#1368), and shipped payout-order discount reads (#1362) and bot auth (#1363). In flight: #1372 x-bot command worker (+24.7K, opened Sat, updated this morning), #555 BagsPay + new fee-share methods (+10.9K, WIP), and bagsapp #692 "parity with web BagsPay" (+18.4K) — mobile is back after two dark weeks (dan also landed 4 WebView/OAuth security fixes, #693). Relay (#541/#1340/#690, +36K) got no update this week; it's cold. Terminal-side nothing shipped; the trading-terminal stock filter (BAGS-412) is still untouched since Sep 14.
Postgres cutover, wave 1, landed. alaa merged the indexer shadow sink (#161, +23.9K, Mon), the reconciler + pg-seed + pg_reconcile worker (#176, +18.1K, Sat) and datapi plain-store family reads with cutover tooling, a kill key and parity CI (#156, +24.3K, Sat), then enabled the Postgres client in production with no read families (automations #142). 109 commits on Saturday alone. Bot fleet to k8s: sadj enabled Dividend and Donation production deployments + Datadog monitors, prepared Cashback (suspended), Airdrop and Compounding (disabled) — 22 automations commits. Security: nikola opened a 7-issue rpc-cache audit (BAGS-489…497, all High), closed 2; Cantina #1346 stale-caches fix merged. Ops incidents: fbmini (the floors + launcher host) lost network twice on Oct 8 (11h and 90min) — the box never rebooted; it's Wi-Fi-only with en0 dead. That outage is where the stock lane stopped. Data: nils shipped top-revenue-tokens (datapi #151) and closed BAGS-492 "DefiLlama evm v3 tracking" — if that adapter lists, the North Star can finally count Robinhood.
Three items that can't wait. (1) The receipt. The Oct 7 tease was the setup; the payoff post (total routed, creators paid, cadence, in SOL, everywhere on earth) is five days late and every day it's late it reads more like edging. The data for it is in our own tables: $36.0K claimed by 121 creators this week, $144K the week before. (2) The hackathon grievance is now inside Finn's replies (LindaSunshine66 "No payment so far!", RyanzoNakamoto "Where is HATON?"). Her own number: 26 winners paid, 7 still building. Pay the outstanding claim, post "all hackathon claims live", and the next official post gets clean replies. (3) Decide the launcher on purpose. It was off all weekend because nobody turned it back on after the Oct 8 outage and Finn's Oct 6 hold. Either it's a product (funded, paced, house wallets tagged) or it stays off and we report organic. Close with named owners.
214 human commits on default branches (176 the week before, +22%), 46 PRs merged (30 by humans), and Linear closed 9. Five humans committed, the same five-person core as last week but a different five: finnbags went from 95 commits to 0 and alaa went from 30 to 142. Saturday Oct 10 alone had 109 commits — the wave-1 Postgres branches landing in one sitting.
Human commits on default branches, Oct 5 → Oct 11 UTC (deploy[bot], bags-deploy-bot, github-actions, renovate and cursoragent excluded). Prior week: finnbags 95, ramyodev 65, unbound-bags 13, nilsbags 2, sadjbags 1 = 176. Linear: 9 closed — 4 audit items (nikola), 4 ops/travel todos (stu), 1 data (nils). nikola's audit work and nils' DefiLlama adapter are the only closes that map to a commit.
| Theme | What landed | Owner | Repo |
|---|---|---|---|
| 🗄 Postgres cutover (wave 1) | Indexer current-state shadow sink turned on (#161/#164), reconciler engine + pg-seed CLI + pg_reconcile worker (#176), datapi plain-store family reads with cutover tooling, readiness floor, kill key and parity CI (#156/#158), Postgres client enabled in production with no read families (automations #142), Datadog distributions (#161). ~66K lines merged; the data behind this deck now has a second store. | alaabags | solana-indexer-v2 · bags.datapi · bags.automations |
| 💸 Claims + BagsPay surface | Token-page claim status + phone layout (#549, finnbags). Claim vault events indexed (#173), claim txs to the claim feed on mobile (#691) and web (#556), claim tweets re-enabled for all claim types (#1368), payout-order discount reads (#1362), bags bot auth (#1363), /api/earnings → claim-stats v4, dev-portal fee-split details. | ramyodev · finnbags | bags.backend · twitter-claim · bagsapp · indexer |
| 🤖 Bot fleet → production GitOps | Dividend + Donation production deployments and Datadog monitors enabled; Cashback deployment prepared (suspended, zero replicas); Airdrop and Compounding prepared (disabled); CNI network policies; dividend-bot release automation, charts, redacted JSON logs, shutdown cleanup. | sadjbags | bags.automations · bags-dividend-bot |
| 🔁 Launch resume | Creators can resume a launch whose fee-share config already landed (#1367, BAGS-487); public-API callers resume with per-wallet allocation compare. Follow-ups open: stalled BagsPay launch resume (#1371), new mint when a stall can't resume (#557, BAGS-488). | unbound-bags · ramyodev | bags.backend · twitter-claim |
| 🛡 Security + audit | Mobile: OAuth state generated and verified on login, WebView token/session injection hardened for wallet migration and export (#693). Backend: stale caches + consistency (#1346). rpc-cache audit opened as 7 High issues (BAGS-489…497), 2 closed, 5 in progress. | danbagsfm · unbound-bags · nikola | bagsapp · bags.backend |
| 📐 Data products | Top revenue tokens endpoint (datapi #151, +1.8K). DefiLlama EVM v3 tracking (BAGS-492). Stock quote catalog refreshed four times (#160/#162/#166/#172). OHLCV live-tail window fixes. | nilsbags · alaabags | bags.datapi · indexer |
Read: infrastructure week. The two things the market can see — claim status on the token page and claim tweets — shipped by Monday and Thursday. The risk moved, not shrank: alaabags + ramyodev = 178 of 214 commits (83%), and Ramy now has three BagsPay PRs open at once (#1372, #555, #557) while owning the payout.
Open PRs went 35 → 39. Three new +10K PRs opened this week, all BagsPay: backend #1372 x-bot command worker (+24.7K), bagsapp #692 parity with web BagsPay (+18.4K), twitter-claim #555 BagsPay + new fee-share methods (+10.9K, WIP). The shadow-store PRs that were "in review" last week all merged. Relay (#541/#1340/#690, +36K) had no commit or comment since before Oct 5. A new contributor, CarloBags, opened #1373 (admin ban endpoint).
| PR | Scope | Owner | Status | |
|---|---|---|---|---|
| HOT | backend #1372 — x-bot command worker | +24,703, opened Sat Oct 11, updated today. A worker that executes X-side commands for BagsPay — the piece that makes "pay via a reply on X" real. Biggest open PR; needs a reviewer who isn't Ramy. | ramyodev | open |
| HOT | bagsapp #692 — parity with web BagsPay | +18,436, opened Oct 8. Mobile gets the BagsPay flows that web shipped Sep 28 – Oct 5. First substantive mobile PR in three weeks. | danbagsfm | open |
| HOT | twitter-claim #555 — BagsPay + new fee-share methods [WIP] | +10,870, updated Oct 9. New fee-share methods on the web launch form; pairs with #1371/#557 (stalled-launch resume, BAGS-488). | ramyodev | WIP |
| HOT | backend #1373 · subgraph #8 · indexer #177 | Admin endpoint to ban pre-generated users (+693, CarloBags, new); v2 hook partner accrual fix on the RH subgraph (+156, nils — relevant to counting RH revenue); Datadog monitor queries (alaa). | CarloBags · nilsbags · alaabags | open |
| COLD | twitter-claim #541 · backend #1340 · bagsapp #690 — Relay cross-chain | +23,385 / +8,937 / +4,083. No activity this week. Last week's call was "hold Relay"; it's being held by neglect rather than decision. Decide: merge date or close. | danbagsfm | idle 7d+ |
| STALE | 5 mobile PRs untouched for 3+ months — oldest bagsapp #602 (Jul 2025) | #602 auth fixes, #609 feed charts, #618 Nativewind (+6.9K), #648 dead-code cleanup, #649 Cashback/Hyperliquid (+25.6K). Unchanged since the Sep 21 deck. | ronickg · danbagsfm · finnbags | 4–15 mo |
indexer (11): #161 shadow sink · #176 reconciler · #173 claim vault events · #169 · #167 · #164 · #171 · 4 catalog refreshes. datapi (6): #156 family reads · #152 shadow client · #151 top revenue tokens · #158 · #160 · #161. backend (7): #1368 claim tweets · #1367 launch resume · #1346 stale caches · #1362 · #1363 · #1360 · #1365. twitter-claim (3): #549 claim status · #556 · #554. bagsapp (2): #693 WebView security · #691. automations (1): #142. Plus 16 release-bot PRs.
Three new +10K BagsPay PRs from two people in one week, on top of three cold +8K–23K Relay PRs. The shadow store cleared because alaa merged his own 66K lines in a weekend; the BagsPay queue won't clear the same way because Ramy is also the payout operator. Sequence: #1372 gets a second reviewer (unbound-bags) this week → #692 → #555. Relay gets a decision, not a fourth week of "in review". 39 open PRs is the highest count since August.
Default-branch commits only, 14 repos. The inversion is one person's branch landing: alaa's shadow-store work was "0 commits, 2 open PRs" last week and "141 commits" this week. bagsapp: 7 commits after two weeks of zero (dan's security fixes). twitter-claim: 10, down from 99 — Finn's launch-form sprint is over.
9 closes: BAGS-489/490 (rpc-cache audit, nikola), BAGS-485/480 (Cantina, nikola), BAGS-492 (DefiLlama EVM tracking, nils), BM-25…28 (stu: stripe, breakpoint, hotel, dinner). 13 created: 7 are nikola's rpc-cache audit (High), plus BAGS-487/488 (launch resume — the first time this quarter a shipped backend feature had a Linear issue before it merged), BAGS-486/491 launchpad coverage, SUPPORT-38. 44 of 55 "started" issues haven't moved in 14+ days (39/49 last week — it got worse). BAGS backlog 119 (14 High, 0 Urgent). The 66K-line Postgres cutover has no Linear issue. Neither does the x-bot worker, mobile BagsPay parity, or the bot-fleet production rollout. The stock-pairs cluster (BAGS-434/436/426/412, High) is now four weeks untouched; we launched 797 stock pairs this week.
Solana did $4.97M vs $17.98M (full week). Excluding high-frequency wallets (≥200 trades in the window), $4.20M vs $12.75M (−67%). The shape: Mon–Tue $230–250K (the pre-$BAGSPAY baseline), Wed $1.86M on the tease, Thu $806K, then $276K–$415K Fri–Sun. Last week's deck asked for Mon–Tue ordinary-wallet volume above $1M to call it a habit; it was $247K and $231K. It was the coin, and then it was the post. Robinhood: 7.07 ETH ≈ $17.7K (−80%), 2 launches.
Ordinary-wallet volume by day: $247K · $231K · $1.86M · $806K · $276K · $415K · $363K. Daily traders: 1.7K · 1.6K · 5.2K · 3.5K · 1.7K · 2.4K · 1.9K. Launching creators: 28 · 70 · 98 · 71 · 30 · 17 · 14. Finn's post went out around midday PT on Wed Oct 7; the UTC day it landed on did 8× the day before in volume, 3× the traders and 1.4× the creators, with zero product change. Half-life ≈ 36 hours: Thu kept 43%, Fri 15%. The receipt that would have made Thu a second spike didn't go out. The lever is proven and it is a post, not a feature. That's the cheapest finding in this deck and the most expensive one to keep ignoring.
Launches by day: 360 · 762 · 323 · 556 · 153 · 19 · 35. Stock pairs: 146 · 336 · 154 · 121 · 15 · 0 · 25. The launches floor was booted out Oct 6 on Finn's hold; the stock lane (300/day from Oct 7) stopped when fbmini lost network on Oct 8 and nobody relaunched it; the brrr.trade copy launcher + backrooms have been off since Oct 6 20:21 PT. 54 farm wallets = 77% of the week's launches, so when our engines stop the counter shows what's underneath: 17 and 14 organic creators a day. Organic for the week: 177 creators, 509 coins (+4%). The 797 stock pairs we did launch Mon–Thu traded $147K total, $184 each. Take rate 0.44% vs DefiLlama by day: $1.3K/$1.6K · $4.0K/$4.6K · $7.8K/$8.0K · $3.3K/$3.3K · $1.2K/$1.2K · $1.9K/$2.0K · $2.2K/$2.2K — still reconciled.
The X API credits are exhausted: x-research and the raw bearer route both return 402 "credits depleted", X syndication 429s, and the web-search fallback provider-errors. Zero fresh tweets were readable on Oct 8–12. Everything below is from the Oct 7 sweep (299 tweets, 24h) plus on-chain behaviour as the proxy for Oct 8–11. Fix is Finn's: top up at console.x.com. Until then this slide is history, not a read.
Net read: Finn still owns the room — resignation flipped to anticipation in eight hours on one image. But a tease has a half-life, the chain says it's ~36 hours, and the follow-up window closed on Oct 9 with nothing posted. The two things that would have compounded it cost nothing: the receipt with numbers (we have them: $36K to 121 creators this week, $144K the week before) and one line that payouts are in SOL, everywhere. Next week's sentiment slide is blank unless the credits are topped up.
The Oct 7 tease is the best A/B test Bags has run: one post, $2.03M, 98 creators, no product change. The pay-off post is five days late and the window has already closed once. Content: total fees routed to creators ($36.0K claimed by 121 creators this week, $144K the week before), the first payout tx, cadence, and the one line that answers the top reply — "payouts in SOL, everywhere on earth; XMoney is the US fiat off-ramp" — without knocking the partner. Everything it links to is live: token-page claim status (#549), claim tweets (#1368), /pay. No CA, no market cap, no competitor names. Owner: Finn (post), Ramy (tx).
Carried from the Oct 7 memo and now visibly more urgent: the grievance thread (HATON / "No payment so far!") colonised the comeback replies within hours. Her own data says 26 winners were paid and 7 still build — that is a retention story we can tell on our terms, but only after the tab is cleared. Cost ≈ 0; earnings exist on bags.fm. Then every official post this week gets clean replies. Owner: ops (stu) + Finn's sign-off.
Weekend launches were 19 and 35 because the launches floor was booted out Oct 6, the stock lane died with the Oct 8 fbmini outage and the copy launcher has been off since Oct 6 — none of it decided, all of it drift. Two honest options. (a) Run it as a product: refund the treasury, pace it, tag house deployers (extend #1354's tag to known wallets) so public stats show organic separately. (b) Leave it off and switch the launch KPI to organic creators (177 this week) and ordinary-wallet volume ($4.2M). Either is fine. Not choosing means next week's counter is whatever fbmini's Wi-Fi did. Also fix fbmini: en0 is dead, the box is Wi-Fi-only, it dropped twice in a day.
797 stock pairs launched Mon–Thu traded $147K, $184 per pair, with 1,232 buyers — four times last week's supply for ten times the volume, still nothing. The product work it needs (BAGS-434 DBC curves, BAGS-436 QA, BAGS-426 dashboard, BAGS-412 terminal filter, all High) is untouched since Sep 14, four weeks now. Keep the lane off until one of those has an owner. Relay (+36K across #541/#1340/#690) had zero activity this week — it's held by neglect. Give it a merge date after the BagsPay queue or close the PRs; a cold +23K PR is review debt, not optionality.
Ramy opened the x-bot command worker (+24.7K) on Saturday, has #555 (+10.9K) WIP and owns the payout. unbound-bags reviews #1372 this week; dan's mobile parity #692 (+18.4K) follows; #555 after. That order lands the thing the receipt post will describe. Separately: nils closed BAGS-492 "DefiLlama EVM v3 tracking" and fixed v2 hook partner accrual on the subgraph (#8). If the DefiLlama adapter lists Robinhood, the $100M counter stops being Solana-only — verify the listing and, if it's live, restate the North Star next week.
Read: Linear tracked an audit and four travel todos. GitHub shipped a database migration. Fourth deck in a row with the same gap.
Not auto-created. Say the word and I'll file these with teams and priorities. Last week's four proposals weren't filed either.
Last week's calendar had Monday "clarity + numbers" and Tuesday "the payout tx". What actually went out was one Wednesday tease from Finn, and it outperformed everything on the plan. So the plan bends to that: Finn's account carries the receipt; @BagsApp carries the proof links. Every post links to something verifiable. No CAs for coins Bags didn't launch, no market caps, no competitor callouts. Every post here is a plan; Finn approves each one.
Cadence rule: TUE (the receipt) is non-negotiable and MON (clear the tab) must precede it. Then a weekly receipt every Monday, so "are they still paying?" always has a dated, public answer. Tease-to-receipt gap: never more than 48 hours again.
Goal: $100M cumulative protocol revenue. Banked all-time: $32.21M (DefiLlama, Solana basis; Robinhood revenue is additive and not included). That's 32.2%. This week added $22.8K (−73% WoW); the on-chain take ($21.8K) matches it, so the number is trustworthy. It is also ~1/110th of the ~$2.5M a week the Q4 target needs with 11½ weeks left. The one day that moved the bar this week was a post, not a product — which is either depressing or the cheapest growth plan Bags has.
| Milestone | Cumulative revenue | Increment | Status |
|---|---|---|---|
| Today (Oct 12) MEASURED | $32.21M | +$22.8K last 7d (Solana public) | 32.2% of goal |
| True incl. RH EST | $32.21M + RH (not re-summed this run) | +≈$177 RH protocol half (7d) | DefiLlama can't see RH — BAGS-492 may change that |
| End of Q3'26 (Sep 30) TARGET | $36M | — | Missed: closed at ~$32.1M |
| End of Q4'26 TARGET | $61M | +$28.8M | needs ~$2.5M/wk · doing $23K/wk |
| $100M GOAL | $100M | +$67.8M | Needs a floor, not a post |
Basis: DefiLlama protocol revenue, Solana (all-time $32.21M, fees $64.43M; 7d $22.8K revenue / $44.8K fees; 30d $167.6K). The honest read: the $BAGSPAY week added 0.08 points; this week added 0.02. The measurement chasm is now tiny in dollars (RH ≈ $177 of protocol take) but structural: a Solana-only counter will under-report the day Robinhood works, and nils' DefiLlama EVM adapter (BAGS-492, closed Thu) is the fix if it lists. Milestone unchanged: $1M/day of ordinary-wallet volume that survives a weekend. This week's best day was $1.86M and it came from a tweet; the floor underneath is $230–415K, ≈$10K/week of revenue at 0.44%. The Q4 target needs 250× that.
bags_trade_feed, UTC days, excluding wallets with ≥200 trades in the two-week window. Including them: Wed 7 = $2.03M and the week = $4.97M. Grey = the week before; red = days at or below the pre-$BAGSPAY baseline. Wed 7 is the tease.
On-chain window scan, chain 4663, exact UTC days Oct 5 → Oct 11 (cache refreshed to block 86.87M, +2 tokens, 4,757 all-time). Week = 7.07 ETH ≈ $17.7K, 2 launches (both Sunday), vs 35.5 ETH / 23 launches the week before (−80%). Faded = the week before. Scale is the prior week's Tue 29 (20.7 ETH).