# ElizaOS Strategic Intel — 2026-01-18

## 1) DATA PATTERN RECOGNITION

### 1.1 Development velocity & trend
- **Repo activity (rolling month-to-date snapshot)**: **32 PRs opened / 18 merged**, **47 issues opened / 35 closed**, **24 active contributors** (elizaos/eliza).
- **Daily execution signal (2026-01-17)**
  - **~10 new PRs submitted** (work-in-progress batch) focused on runtime/CLI correctness and stability.
  - **Core runtime enhancements completed**:
    - Added `unregisterAction` (improves action lifecycle flexibility).
    - Fixed auth/.env loading in agent commands; CLI directory validation; MESSAGE_SENT event emission correctness; plugin-bootstrap fact persistence connection ordering.
  - **Dependency churn**: multiple core + dev dependency upgrades (zod, react-markdown, tailwind-merge, p-retry, @types/node, prettier, etc.). Trend indicates active modernization but increases regression surface area.

**Trend callout:** Engineering effort is clustering around **runtime correctness + infrastructure hardening**, while community discussions cluster around **token migration clarity** and **Polymarket integration reliability**. This is a healthy “stabilize-the-core” phase, but user-perceived value is being capped by unresolved “sharp edges” (migration UX, plugin-discord breakage, embedding config mismatch).

### 1.2 Community engagement patterns (Discord)
- High-engagement threads concentrated in:
  1) **Token economics / migration** (deadline pressure, consequences for non-migration unanswered).
  2) **Trading/prediction integrations** (Polymarket wallet model + Cloudflare/egress restrictions).
  3) **Jeju network vision** (autonomous agents, secure containers, ICP runtime success).
- **Support interactions** are primarily peer-to-peer (sayonara, lucky_beagle_52756, Chucknorris), indicating strong community “ops,” but also that official docs/product guardrails are insufficient for recurring issues.

### 1.3 Feature adoption & usage signals
- **Polymarket plugin/example**: active experimentation; friction concentrated in operational details (Safe proxy wallet flow + serverless egress blocks).
- **Docs AI tool**: strong positive testimonial (“6 hours > 2 months”), suggesting meaningful adoption and high ROI for improving discoverability/onboarding.

### 1.4 Pain point correlation across channels
| Pain Point | Where it shows | Likely root cause | Impact |
|---|---|---|---|
| Embedding dimension mismatch / USE_OPENAI_EMBEDDING ignored (Issue #6380) | Engineering triage + runtime failures | Config precedence + adapter assumptions | **High** (hard failures, SQL errors) |
| Discord integration error (`sendMessage` undefined) | Discord user report + urgent triage note | Version skew between plugin-discord + CLI/core | **High** (breaks a major channel) |
| Polymarket: wallet address mismatch | Coders channel | Safe proxy architecture not reflected in mental model/docs | **Medium-High** (blocks trading use cases) |
| Polymarket: Cloudflare 403 in serverless | Coders channel | Egress/IP reputation + runtime restrictions (Supabase Edge) | **Medium-High** (blocks deployment archetype) |
| Token migration confusion (deadline + consequences) | Discussion channel | Inconsistent bot messaging + missing policy docs | **High** (trust + conversion risk) |
| “Why would agent creators reward holders?” | Discussion channel | Missing incentive design narrative | **Medium** (token utility skepticism) |

---

## 2) USER EXPERIENCE INTELLIGENCE

### 2.1 Feedback categorized by impact & theme

**P0 (Breaks core usage)**
- **Discord:** `this.runtime.elizaOS.sendMessage is undefined` blocks Discord messaging.
- **Embeddings/DB:** `USE_OPENAI_EMBEDDING` not respected causing SQL dimension mismatch errors (Issue #6380).

**P1 (Blocks key integrations / monetizable workflows)**
- **Polymarket wallet model confusion:** Safe (1/1) proxy wallet vs EOA mismatch; exported private key not matching “Builder address.”
- **Polymarket deployment friction:** Cloudflare 403 from Supabase Edge Functions; proxy and custom CA limitations.

**P2 (Conversion, trust, and narrative)**
- **Token migration unanswered policy question:** what happens to unmigrated tokens (lost vs burned vs redistributed) remains unanswered.
- **Incentive design gap:** unclear value exchange for “agent creation rewards holders.”
- **Prompt UX constraint:** request to remove **500-character limit** in first application prompt (Issue #6381) suggests onboarding prompt is constraining real use.

### 2.2 Usage patterns vs intended design
- **Intended:** “Example plugin works; deploy anywhere; agents integrate with markets.”  
  **Observed:** users choose serverless (Supabase Edge) for speed-to-deploy, but **edge runtime constraints** (no proxy env, no custom CA) conflict with real-world APIs protected by Cloudflare and with compliance/IP heuristics.
- **Intended:** plugin abstracts Polymarket auth.  
  **Observed:** users must understand Safe proxy flow to debug; abstraction leaks without docs and logging.

### 2.3 Implementation opportunities (high leverage)
- Add **first-class “Integration Runbooks”** for:
  - Polymarket: Safe proxy explanation + required headers + troubleshooting address mismatch.
  - Cloudflare-protected APIs: guidance on dedicated egress IPs, non-edge deployment template, compliance checklist.
- Improve **configuration validation** at startup:
  - Detect embedding dimension mismatch early; fail with actionable remediation (expected dims, adapter dims, model dims).
- Fix onboarding friction:
  - Remove/raise initial prompt char limit or make it model/context-aware (Issue #6381).

### 2.4 Community sentiment (qualitative)
- **Positive:** excitement about ICP runtime success; strong endorsement for docs AI tool; active builder energy in coders channel.
- **Negative/at-risk:** token chart anxiety, migration confusion, skepticism about prediction market hype, and incentive design doubts.

---

## 3) STRATEGIC PRIORITIZATION (Impact × Risk × Dependency)

### 3.1 Priority stack (next 7–14 days)

#### P0 — Reliability Hotfix Track (stop the bleeding)
1) **Fix Issue #6380 (Embedding config mismatch)**
   - **User impact:** prevents runtime SQL failures; blocks agent startup for affected configs.
   - **Technical risk:** medium (touches config precedence + DB adapter assumptions).
   - **Recommendation (implementation):**
     - Enforce a single source of truth for embedding dimension: `EMBEDDING_DIMENSION` (character) > adapter introspection > provider call.
     - Add startup assertion: if stored embeddings table dimension ≠ current model dimension, output migration instructions or auto-migrate (if safe).
2) **Resolve Discord sendMessage undefined (version skew)**
   - **User impact:** breaks a core comms channel.
   - **Dependency:** plugin-discord PR + CLI version alignment (**@elizaos/cli 1.7.1 noted**).
   - **Recommendation:**
     - Publish a compatibility matrix: core/cli/plugin-discord versions.
     - Add runtime guard: detect missing `sendMessage` capability and degrade gracefully with a clear error.

#### P1 — Integration Enablement Track (unblock builders)
3) **Polymarket “Safe proxy wallet” as a first-class concept**
   - **User impact:** reduces time-to-first-trade; reduces funds-at-risk mistakes.
   - **Recommendation:**
     - Add plugin-level logging: resolved signer address, proxy address, and active trading address (redacted).
     - Add docs: “Exported PK corresponds to signer EOA; funds live in Safe proxy; interact via proxy.”
4) **Cloudflare 403 in serverless: formalize supported deployment archetypes**
   - **User impact:** reduces dead-end architecture choices.
   - **Recommendation:**
     - Officially recommend **non-edge** execution for Polymarket CLOB (VM/container) with **static egress IP**.
     - Provide a reference deployment: Docker + Postgres + worker queue (BullMQ/NATS depending on throughput).
     - Document why Supabase Edge fails (no custom CA / proxy env) and offer alternatives.

#### P2 — Product/Narrative Track (trust + conversion)
5) **Token migration policy clarification**
   - **User impact:** reduces uncertainty; improves trust.
   - **Recommendation:** publish an official policy note answering: “unmigrated tokens by deadline: lost/burned/redistributed?” and pin it; align bot messaging with announcements.
6) **Incentive design: “why would creators reward holders?”**
   - **User impact:** reduces skepticism; improves token utility narrative.
   - **Recommendation:** produce 2–3 viable incentive models with abuse constraints (see §3.3).

### 3.2 Resource allocation recommendation (2-week view)
- **60% Core Reliability (P0):** embeddings config + Discord version skew + regression tests around config precedence and plugin API compatibility.
- **30% Builder Enablement (P1):** Polymarket runbook + deployment template + improved plugin diagnostics.
- **10% Narrative/Docs (P2):** migration policy + incentive design explainer; ensure docs are discoverable from CLI errors and README entry points.

### 3.3 Decision-ready proposals

#### A) Holder rewards tied to agent creation — viable incentive structures
To answer the core objection (“why would creators pay holders?”), only proceed if it’s **creator-neutral** or **creator-positive**:
1) **Protocol fee rebate model (creator-neutral):**
   - Agent creation includes a network fee paid by user; portion routes to holders via staking pool.
   - Creator doesn’t pay; users pay for provisioning/hosting.
2) **Growth rebate model (creator-positive):**
   - Creator gets credits/reduced hosting cost proportional to their agents’ usage; the rebate is funded by protocol revenue; holders share in revenue.
3) **Opt-in patronage model (market-driven):**
   - Creators can opt into “support holders” badge that increases discovery ranking; a small voluntary skim funds holders.

**Abuse prevention baseline:** sybil-resistant eligibility, minimum stake duration, rate limits per creator/user, and agent creation fee floor.

#### B) Jeju / autonomous agent infra — critical path dependencies
- DorianD’s requirements imply three deliverables that must be sequenced:
  1) **Provable provisioning** (secure containers / enclaves, attestation story)
  2) **Agent registration + identity**
  3) **Protocol support (8004, 402 / x402 payments)**
- Recommendation: publish a **Jeju MVP scope** explicitly optimized for **async/long-running tasks** (aligning with earlier latency concerns) + an “oracle posting” reference agent to demonstrate end-to-end value.

---

## 4) METRICS TO WATCH (next report)
- **Time-to-first-success** for Polymarket example (clone → funded → first order) and top failure reasons.
- **Incidence** of embedding dimension errors post-fix (Issue #6380).
- **Discord plugin health**: message send success rate; version mismatch occurrences.
- **Migration support load**: # of repeated questions/day; unresolved FAQ count (target: drive to near-zero with pinned policy + bot alignment).
- **Doc AI tool adoption**: clicks/usage and correlation with reduced support requests.

---

## 5) ACTIONABLE NEXT STEPS (owner-ready)

### Engineering (P0)
- Ship hotfix for **USE_OPENAI_EMBEDDING / embedding dimension** mismatch with startup validation and remediation output.
- Resolve **plugin-discord + CLI** compatibility; publish version matrix; add guardrails.

### Developer Experience (P1)
- Publish “Polymarket Integration Runbook” (Safe proxy wallet, headers, debugging).
- Provide non-edge deployment template with static egress IP guidance (explicitly discourage Supabase Edge for this integration).

### Product/Comms (P2)
- Publish token migration consequence policy and pin it; synchronize migration bot messages with official announcements.
- Draft incentive design memo for holder rewards with at least one creator-neutral model and explicit anti-abuse rules.