SaySo — State of the Project

TL;DR

SaySo (Oracle) is in pre-launch alpha — the platform is ~90% built and the team has aligned on shipping over building (see: apps/pm-v2/strategy/meeting-notes-2026-04-07.md). Over the last weekend JP landed survey gating + UX fixes on the app (8+ participant auto-survey, reworded Q3, email alerts, join-link bug fix) (see: sayso/meetings/2025-05-11-sayso-okr-summary.md). The platform has had one real-world session so far (run by streamer Paul's teammate); zero organic sessions. The week's primary GTM bet has shifted: pitching Top Shot on a non-stream SaySo-powered event is now the active concierge play (JP owns; waiting on Matt + Top Shot reply). The 2026-04-07 "DM 30 Polymarket bot builders" item was not raised this week — flagged as drifted, not formally killed. Active dev is still on the oracle-stag branch (no new commits since 2026-04-27; uncommitted work remains).

Immediate next steps

The latest meeting is the 2026-05-11 OKR sync (lives in wiki: sayso/meetings/2025-05-11-sayso-okr-summary.md). It drives the top of this list. Older 2026-04-07 items that weren't re-affirmed are flagged below. Items already shipped via code or via the weekend SaySo-app shipments are excluded.

Drifted (not re-affirmed in latest sync)

DM 30 Polymarket bot builders + 20 agent developers (concierge onboarding from 2026-04-07, owner Gui) was not discussed at the 2026-05-11 sync. Top Shot pitch has effectively replaced it as this week's concierge bet. Not formally killed — flag for operator: pause or kill? (from: apps/pm-v2/strategy/meeting-notes-2026-04-07.md vs. sayso/meetings/2025-05-11-sayso-okr-summary.md)
Adding "you can host this yourself" copy to the survey was suggested by Simon but rejected by Ben — surveys must stay concise. Implied follow-up: this becomes an onboarding/discovery question for the product, not a survey question. (from: sayso/meetings/2025-05-11-sayso-okr-summary.md, 2026-05-11)

Phase / Milestone progress

Phase: Pre-launch alpha — "platform mostly built, ship and validate" (see: apps/pm-v2/strategy/launch-plan-summary.md)
Real-world usage: 🟡 ~1 session so far (run by a teammate of streamer Paul); 0 organic (see: sayso/meetings/2025-05-11-sayso-okr-summary.md, 2026-05-11)
Survey infrastructure: ✅ shipped weekend of 2026-05-09/10 — auto-trigger at 8+ participants, Q3 reworded ("what could have made this better?"), email alerts, join-link bug fixed (from: JP shipment list in sayso/meetings/2025-05-11-sayso-okr-summary.md; not yet visible in git log — likely on oracle-stag working tree or a side branch)
Score Points + Tiers + XP: ✅ shipped (migrations 056–063, tier model Script→Bot→Drone→Synth→Cyborg→Sentinel→Sovereign→Overlord→Prime) (commit e278026f, 2026-04-13)
User-type differentiation (humans vs agents): ✅ shipped (PR #286) — closes the "Differentiate bots from humans" TODO (commit e278026f, 2026-04-13)
Prediction-market creator flow: ✅ shipped (PR #282) (commit 46b8929c, 2026-03-18)
WebSocket support + replay hydration: ✅ shipped (PR #283 + follow-up) (commits 77616b72, 6d2aff04)
Agent event system + heartbeat: ✅ shipped (commits 655bf8b5, 1d82274e)
MCP integration path: ✅ packages/sayso-mcp-server v1 + sayso-claude-plugin/skills shipped (alongside the universal skill.md fallback) (commit 8c50b773, 2026-04-21)
Batch trading + user positions: ✅ shipped (commit 1f99bb89, 2026-04-22)
better-auth provider migration: 🟡 in flight on oracle-stag (migration 067 staged but uncommitted)
Top Shot event pitch: 🟡 in flight — JP awaiting reply from Matt (and Guy/Gee on Top Shot side); branding-off-the-surface hypothesis from Naim suggests NBA may not be required (see: sayso/meetings/2025-05-11-sayso-okr-summary.md, 2026-05-11)
Season 1 design: 🟡 spec approved (V7 SP formula validated by Monte Carlo) — launch decision still pending (see: apps/pm-v2/strategy/competition-design.md)
Retention instrumentation: 🔴 not yet built — this is THE metric per launch plan; moot until volume exists per 2026-05-11 (see: apps/pm-v2/strategy/launch-plan-summary.md)
Real-money path for Season 2: 🔴 not started; gate at Week 5 of S1 (see: apps/pm-v2/strategy/launch-plan-summary.md)

Roadmap

This week (2026-05-11 → 2026-05-17)

Run the Top Shot pitch flow end to end (poke Matt → confirm format → casual or formal pitch) (see: sayso/meetings/2025-05-11-sayso-okr-summary.md)
JP ports Jad's sandcastle pipeline to SaySo for small-wins iteration cadence (see: sayso/meetings/2025-05-11-sayso-okr-summary.md)
Ben out Wed 2026-05-13 (PTO) — schedule around it (see: sayso/meetings/2025-05-11-sayso-okr-summary.md)
Capacity drain alert: Simon pivots to the AZ website refresh today (Sam's urgent ask); affects SaySo bandwidth this week

Next 2 weeks

JP off 2026-05-18 → 2026-05-22 (cabin at Watcom — laptop "just in case"). Any Top Shot pitch follow-through needs a clear handoff before this window (see: sayso/meetings/2025-05-11-sayso-okr-summary.md)
Land the better-auth provider migration + MCP handler refactors currently uncommitted on oracle-stag (branch: oracle-stag)
Lock Season-1 format, prize, and seeded-bots policy — answers from team meeting still required (see: apps/pm-v2/strategy/meeting-notes-2026-04-07.md)
Stand up retention dashboards (Day-7 / Day-14 agent return) (see: apps/pm-v2/strategy/launch-plan-summary.md)
Begin the daily X ritual (3 replies + 1 post + 1 DM) and audit follows / build 3 X lists (see: sayso/notes/marketing-research/sayso-x-todo.md)

Next month

Launch Season 1 (Week 1 of launch plan): AI creators + AI players go live, X account active, talk to Flow team (see: apps/pm-v2/strategy/launch-plan-summary.md)
Ship the leaderboard daemon (@SaySoArena auto-posts) — primary content engine (see: sayso/notes/marketing-research/sayso-x-marketing-playbook.md)
Submit to MCP registries; post in LangChain / CrewAI Discords (see: apps/pm-v2/strategy/launch-plan-summary.md)
Mid-season Week-5 gate check: 5 churn interviews + 5 retention interviews + WTP signal (see: apps/pm-v2/strategy/launch-plan-summary.md)
Jad's 4-day Eid PTO lands end of month — plan around it

Next quarter / beyond

Season-1 → Season-2 decision: accelerate / adjust / kill at Week 10 based on Day-14 return rate (>20% accelerate, 10–20% adjust, <10% kill or pivot) (see: apps/pm-v2/strategy/launch-plan-summary.md)
If accelerating: real-money mechanism, Product Hunt launch, framework integrations (LangChain / CrewAI / Claude Code plugin registry) (see: apps/pm-v2/strategy/launch-plan-summary.md)
Long-term thesis: Agent Reputation Protocol — portable Elo / win-rate / accuracy as a credential before any company deploys an agent with real capital. PMs are the first benchmark; the data IS the moat. (see: apps/pm-v2/strategy/growth-plan.md; apps/pm-v2/strategy/bet-proposal.md)
Marketing flywheel — adopt newswire formatting (Polymarket pattern), one weekly partnership, Farcaster mini-app embed (see: sayso/notes/marketing-research/prediction-markets-solo-playbook.md; sayso/notes/marketing-research/predict-fun-marketing-playbook.md)

Recent shipments

Last 14 days (since 2026-04-27).

Survey gating removed — auto-trigger on any session with 8+ participants (previously gated to JP) (JP, weekend 2026-05-09/10; reported in sayso/meetings/2025-05-11-sayso-okr-summary.md)
Survey Q3 rewritten — from "was anything off about this session?" to "what could have made this better?" (JP; same source)
Email alert wired up — JP pinged whenever a session qualifies for a survey (JP; same source)
Join-link bug fixed — new users following a join link no longer get bounced into the join-code fallback (JP; same source)
Add full snapshot + TODO documentation files (consolidated current backlog and project snapshot) (commit ebf6f369, 2026-04-27)

Last 30 days (broader sprint context):

MCAP API batch trading + user positions (mcp_batch.go, position endpoints) (commit 1f99bb89, 2026-04-22)
Agent event + heartbeat features (live agent presence signaling) (commit 1d82274e, 2026-04-22)
Social and profile features for the MCP surface (commit cb519bf3, 2026-04-21)
MCP integration & user-context — packages/sayso-mcp-server v1 + sayso-claude-plugin/skills/sayso/ (commit 8c50b773, 2026-04-21)
Provider config support (multi-provider readiness) (commit da6062ad, 2026-04-20)
WebSocket replay hydration for prediction markets (commit 6d2aff04, 2026-04-20)
Agent event system (events table + worker + DTOs) (commit 655bf8b5, 2026-04-20)
User comments search + reactions polish (commits f7adddea, 78141b62)
User-type differentiation across MCAP API + PM v2 (PR #286) — bots vs humans across leaderboard, XP, score points, tiers (commit e278026f, 2026-04-13)

Key decisions to date

Agent-first positioning — SaySo treats AI agents as first-class participants; humans are spectators / co-traders. Two-minute onboarding via skill.md. (see: apps/pm-v2/strategy/brand.md; apps/pm-v2/strategy/product-soul.md)
Play money for validation, real money to follow within ~60 days (Board Advisory 9-1) (see: apps/pm-v2/strategy/growth-plan.md)
Competition-driven loop, not open-ended platform — Season 1 = 2 months, Elo / SP-ranked (see: apps/pm-v2/strategy/launch-plan-summary.md; apps/pm-v2/strategy/competition-design.md)
Score Points V7 formula validated by 1000-run Monte Carlo across 7 archetypes; bet size does not affect SP (skill = prediction accuracy, not capital management) (see: apps/pm-v2/strategy/competition-design.md)
Tier ladder: Script → Bot → Drone → Synth → Cyborg → Sentinel → Sovereign → Overlord → Prime; starting SP = 500 (Bot) (see: apps/pm-v2/strategy/competition-design.md)
$100/day credits, no balance cap for M1 (see: apps/pm-v2/strategy/competition-design.md)
Headpin ICP = Polymarket bot builders (200–500 active devs) — not "anyone with an agent" (see: apps/pm-v2/strategy/bet-proposal.md)
Kill criterion: Day-14 agent return rate < 10% after 30+ agents — but moot until session volume exists, per 2026-05-11 (see: apps/pm-v2/strategy/launch-plan-summary.md; sayso/meetings/2025-05-11-sayso-okr-summary.md)
MCP server is the premium agent integration path; skill.md remains the universal fallback (committed as packages/sayso-mcp-server)
X is the primary marketing channel; build a leaderboard daemon as the core content engine (see: sayso/notes/marketing-research/sayso-x-marketing-playbook.md)
Surveys stay concise — host-yourself discovery is a product/onboarding fix, not a survey fix (decided 2026-05-11; see: sayso/meetings/2025-05-11-sayso-okr-summary.md)

Open questions / blockers

Real-world usage is still effectively zero. Only one session has run (streamer-led); the survey infrastructure now exists but is only valuable once sessions happen. The Top Shot pitch is the bet on producing real volume. (see: sayso/meetings/2025-05-11-sayso-okr-summary.md, 2026-05-11)
NBA approval risk on Top Shot event. Top Shot indicated NBA may be a blocker; Ben's hypothesis is that keeping SaySo branding off the surface sidesteps NBA involvement. Unconfirmed; needs Top Shot's read once they reply. (see: sayso/meetings/2025-05-11-sayso-okr-summary.md)
Onboarding gap — users invited to a session may not realize they can host their own. Explicitly deferred from the survey channel; needs a product/onboarding answer. (see: sayso/meetings/2025-05-11-sayso-okr-summary.md)
DM 30 / DM 50 plan status unclear. Item from 2026-04-07 was not raised at the 2026-05-11 sync. Is it paused while Top Shot is the focus, or formally killed? Operator decision needed.
Survey shipments aren't in git log. JP described 4 weekend shipments (gate removal, Q3 copy, email alert, join-link fix) but the public main branch sits at commit ebf6f369 from 2026-04-27. Likely on oracle-stag working tree, a side branch, or a runtime config. Worth confirming the source-of-truth state.
Filename date typo. sayso/meetings/2025-05-11-sayso-okr-summary.md and …-transcript.md use a 2025 prefix even though the meeting was 2026-05-11. Cosmetic but breaks reverse-chrono sorting against future meetings.
Capacity drain on SaySo this week. Sam pulled Simon onto the AZ website refresh; Ben + Jad shipping cross-project agent/pipeline infra (Toby first). SaySo time this week is largely JP solo.
PTO blocking forward motion. Ben out 2026-05-13, JP out 2026-05-18 → 2026-05-22, Jad 4-day Eid end of month. Any Top Shot pitch follow-through after the 17th needs a handoff to Ben or Gui.
No SaySo meetings lived in the wiki before this run. Fixed today — sayso/meetings/2025-05-11-sayso-okr-summary.md is the first. Recommend mirroring all future SaySo syncs into sayso/meetings/YYYY-MM-DD-<topic>.md.
Wiki path nesting. Existing SaySo docs sit at sayso/notes/marketing-research/<doc>.md (4-level), and meetings now live at sayso/meetings/<doc>.md (3-level). The soul's rules state "wiki is 2-level only" — but the underlying system clearly accepts deeper paths. Flag for operator: accept the deeper structure or flatten?
TODO.md drift. TODO.md has uncommitted additions (mini-apps for NBA / soccer, sweepstakes, quick markets, "Plan → Trigger Skill Agent") that have no corresponding code or strategy doc. Decision needed: which become real next-cycle work?
oracle-stag is the active branch with substantial uncommitted state — better-auth migration 067, MCP handler refactors, lottery price worker tweaks, weekend survey/UX work. Risk of drift if not landed soon.
Seeded-bots visibility on the leaderboard — flagged as JP-owned in 2026-04-07; no resolution surfaced 2026-05-11.
Legal check on cash prizes with play tokens — Gui flagged "thinks yes, needs confirmation" 2026-04-07; unresolved.
Real-money mechanism for Season 2 — design has not started; required gate per launch plan.
Marketing channels are documented but inactive. apps/pm-v2/GTM-CLAUDE.md defines the content/ workflow; the content/twitter/published/ and other folders are empty (.gitkeep only). The X ritual hasn't started.
Apps not yet documented in the wiki — apps/pm-hoops, apps/social-pm-app, apps/social-pm-flow, apps/predictions-market-app, apps/azsavings-*, apps/crash_game, apps/gem-slot, apps/game, apps/mcap-on-chain, apps/mcap-games-api, apps/ck-mcap. Open question: which are SaySo-relevant and should be tracked here?
predict-fun-marketing-playbook.md and prediction-markets-solo-playbook.md are project-agnostic playbooks living in sayso/. Acceptable, but operator may prefer they live in a shared marketing/ section.
Risk Jad raised (2026-04-07) is unresolved: what stops Polymarket from copying the agent-arena wedge in a week? The strategic answer ("structural disruption asymmetry, data moat") is in apps/pm-v2/strategy/growth-plan.md but hasn't been re-litigated with the team.

Doc index

SaySo — State of the Project — this dashboard (the live operating view)
SaySo OKR Sync — 2026-05-11 (Summary) — latest team sync; SaySo decisions + non-SaySo capacity context
SaySo OKR Sync — 2026-05-11 (Transcript) — raw transcript backing the summary
SaySo / PM Apps — X Marketing Playbook — daily X ritual, AI-assisted compression, asset roadmap, tool stack, who-to-study
SaySo X — Starter TODO — checklist companion to the playbook
predict.fun — Marketing Strategy Reverse-Engineered — 8-pattern teardown of predict.fun's rise
Prediction Markets — Solo-Founder Marketing Playbook — cross-platform synthesis with a 90-day solo plan

Codebase strategy docs (not in wiki — referenced for provenance)

apps/pm-v2/strategy/product-soul.md — Socratic product-soul extraction (2026-04-06)
apps/pm-v2/strategy/brand.md — brand positioning, voice, audience, competitors
apps/pm-v2/strategy/growth-plan.md v2 — growth-loop classification, AARRR, network effects
apps/pm-v2/strategy/competition-design.md — SP V7, tiers, season mechanics
apps/pm-v2/strategy/launch-plan-summary.md — week-by-week launch with decision thresholds
apps/pm-v2/strategy/bet-proposal.md — formal bet proposal with JTBD + four forces
apps/pm-v2/strategy/meeting-notes-2026-04-07.md — prior team sync (Gui, Ben, JP, Jad, Sam); superseded for near-term steps by 2026-05-11 sync
apps/pm-v2/GTM-CLAUDE.md — content workflow + naming conventions
TODO.md — running backlog (uncommitted additions present)
full-snapshot.md — captured Oracle UI / structural snapshot

SaySo — State of the Project

TL;DR#

Immediate next steps#

Drifted (not re-affirmed in latest sync)#

Phase / Milestone progress#

Roadmap#

This week (2026-05-11 → 2026-05-17)#

Next 2 weeks#

Next month#

Next quarter / beyond#

Recent shipments#

Key decisions to date#

Open questions / blockers#

Doc index#

Codebase strategy docs (not in wiki — referenced for provenance)#