# ElizaOS Intel — 2026-02-25

## 1) Data Pattern Recognition (Velocity, Engagement, Adoption, Pain Correlation)

### Development velocity (GitHub, Feb-to-date snapshot)
- **PRs:** 38 opened / **18 merged** (47% merge rate)
- **Issues:** **41 opened** / **65 closed** (net issue burn-down)
- **Contributors:** **35 active**
- **Code churn (reported):** **+18,576 / -3,807** across **160 files**, **140 commits**
- **Signal:** Despite Discord “low activity” days, engineering throughput is materially active; perception gap is primarily *comms and product packaging*, not lack of code.

### Community engagement patterns (Discord, last 72h)
- **Topic concentration:** token price/viability reassurance, revenue pivot clarification, and admin/access friction.
- **Engagement quality:** low technical depth in public channels; most “builder energy” is either in core-dev discussions (restructure) or external tool/plugin announcements.
- **Sentiment trend:** fragile/conflicted—users see shipping signals (hackathon submission, new tool) but still anchor on **price-to-activity disconnect**.

### Feature/tool adoption signals
- **plugin usage guidance:** community is being redirected from `plugin-image-generation` → **`plugin-bootstrap` image generation action** (implicit consolidation pattern).
- **New external tool:** **fomolt** introduced as an agentic trading test platform (Base, CLI, paymaster, API keys). This is a strong developer-experience lever if we integrate/document it.

### Pain point correlation across channels (recurring themes)
1. **Token/cost & context bloat**
   - Discord: 200k token limit issues attributed to providers/evaluations and message/reflection accumulation.
   - GitHub: open bug **URL triggers duplicate LLM calls** → direct cost + UX regression.
2. **Migration + versioning confusion**
   - v2 DB refactor removes 1.4→1.6 auto-migration; implied assumption users are “already migrated,” but community still shows upgrade uncertainty/scam concerns.
3. **Access/permissions friction**
   - Manual Discord role/channel access needed even for core contributors → slows coordination and reinforces “project is stalled” narratives.

---

## 2) User Experience Intelligence (Feedback, Usage vs Design, Opportunities, Sentiment)

### Feedback categorized by impact × theme

**High impact**
- **Cost/Token efficiency regressions**
  - *Symptom:* duplicate LLM calls on URL messages; frequent token-limit hits with many plugins.
  - *Impact:* direct cost blow-ups, degraded response quality, slower runs.
- **Trust/safety confusion (migration + scams)**
  - *Symptom:* users unsure about token migration legitimacy; warned about DM scams.
  - *Impact:* churn risk + support load + reputational damage.

**Medium impact**
- **Onboarding + monetization readiness**
  - *Symptom:* Eliza App “needs onboarding completion to begin charging”; Milady positioned as funnel to Cloud.
  - *Impact:* delays revenue capture and weakens narrative during bear market.
- **Admin friction**
  - *Symptom:* role assignment limitations; manual channel adds required.
  - *Impact:* slows new dev ramp, creates “closed/opaque” perception.

**Low impact (but sentiment-relevant)**
- **Price discussion displacement**
  - *Symptom:* repeated “project dead?” queries; routed to trading channel.
  - *Impact:* ongoing morale drain unless countered by consistent shipping + metrics.

### Usage patterns vs intended design
- **Users are pushing plugin-heavy agents** (many providers/evaluators) beyond baseline assumptions, causing prompt bloat and token ceiling hits.
- **Developers want safer “test trading” loops** (fomolt) without seed phrases—aligns with intended agent experimentation, but needs first-party patterns/examples to become standard.

### Implementation opportunities (near-term)
- **“Cost & context guardrails” as default UX**
  - Add/enable scenario evaluators and runtime protections to prevent duplicate calls and runaway context growth.
- **One canonical “Trading Agent Test Harness”**
  - Standardize a reference integration for fomolt (or equivalent) so community can reproduce quickly and safely.
- **Migration clarity pack**
  - One page: “v1.x → 1.6.x → v2.0.0” with explicit errors/messages and anti-scam guidance.

### Community sentiment (what to message back)
- Reassurance alone is no longer sufficient; community is asking for **evidence artifacts**:
  - weekly ship notes, revenue product milestones, and clear “what to try today” developer pathways.

---

## 3) Strategic Prioritization (Impact vs Risk, Dependencies, Resource Allocation)

### Current strategic context (from core-dev)
- Organization restructuring toward **small venture teams** with fixed budgets and rapid kill/ship decisions.
- Revenue projects in flight:
  - **Hyperscape Agent Arena duel betting** (launch next week; 2%+1% fees)
  - **OTC agent @ tradingdesk.fun** (1% fee)
  - **Milady** (FOSS funnel → Eliza Cloud)
  - **Babylon** (1% fees; ramp 1%→100% rollout)
  - **Eliza App** (needs onboarding+billing to charge)
  - **Amelia subscription project**

### Critical path dependencies (most likely to bottleneck)
1. **Milady ship → Cloud conversion loop**
   - Without Milady live and instrumented, Cloud revenue thesis is unproven.
2. **Eliza App onboarding + billing**
   - Monetization blocked until onboarding and billing are complete and reliable.
3. **v2 DB refactor + migration path**
   - If migration errors are unclear, support/scam fears increase and adoption stalls.
4. **Token efficiency fixes**
   - If token blow-ups persist, power users (plugin-heavy) become negative advocates.

### Initiative scoring (Impact / Technical Risk / Time-to-value)
| Initiative | User Impact | Tech Risk | Time-to-value | Recommendation |
|---|---:|---:|---:|---|
| Fix **URL duplicate LLM calls** | High | Low-Med | Fast | **Do now (hotfix)** |
| Reduce **prompt/context bloat** (providers/reflections) | High | Med | Medium | **Top 3 engineering focus** |
| Publish **Migration clarity + anti-scam** docs | High | Low | Fast | **Do now (owner: DevRel/Eng)** |
| Eliza App **onboarding + billing** | High (revenue) | Med | Medium | **Staff aggressively** |
| Milady **ship + instrument funnel** | High (revenue) | Med | Medium | **Must-not-slip milestone** |
| fomolt **reference integration** | Medium | Low | Fast | **Do this week** |
| Discord **role automation / access playbook** | Medium | Low | Fast | **Quick ops win** |

### Resource allocation (practical)
- **Core Eng (2–3 ppl equivalent):** token efficiency + URL duplication + migration guardrails (directly reduces cost + support).
- **Product/Revenue squad(s):** Milady ship + Eliza App onboarding/billing (revenue unlock).
- **DevRel/Community (1 owner):** weekly ship/revenue scoreboard + “What to try” posts + anti-scam canonical link hub.
- **Ops/Moderation (part-time):** Discord permissions automation and onboarding checklist for new devs.

---

## Quantitative Recommendations (what to measure next)
1. **Cost & efficiency KPIs (weekly)**
   - Median tokens per response (by agent template)
   - % messages hitting token limits
   - Duplicate-call incidence (URL case) after fix
2. **Revenue readiness KPIs**
   - Eliza App onboarding completion rate → first paid conversion
   - Milady funnel: installs → Cloud signups → retained weekly actives
3. **Community health KPIs**
   - # of “project dead/price” concerns per day (should fall as ship notes increase)
   - Time-to-access for new contributor roles (target <24h)

---

## Actionable Next Steps (next 7 days)
1. **Ship a hotfix** for the webapp bug: URLs triggering duplicate LLM calls (cost + UX critical).
2. **Create a “Context Budget” default**:
   - cap recentMessages/reflections payload, summarize older reflections, and/or make providers lazy/conditional.
3. **Publish a single migration truth source**:
   - “If you’re on 1.4.x, upgrade to 1.6.x first; v2 will error with clear instruction.” Include scam warnings and official links only.
4. **Turn fomolt into an official example**:
   - one repo/example agent, minimal keys flow, Base paymaster assumptions, and a safety note (limits, dry-run mode).
5. **Operational fix:** Discord role escalation/runbook (and document channel IDs / access steps) to eliminate manual troubleshooting loops.
6. **Weekly public scoreboard** (Discord + X):
   - shipped items, revenue product progress, and “try it today” links (reduces viability FUD without debating price).