# ElizaOS Intel — 2026-02-15

## Executive Signals (last 72h)
- **Operational friction surfaced in two places**: (1) *token migration window closed* → users who missed deadline have no path; (2) *cloud credits managed via direct DB UPDATE* → works but is non-scalable and audit-risky.
- **Product positioning gap is now the dominant UX risk**: repeated questions on “non-technical onboarding,” roadmap clarity, and token utility indicate expectation mismatch (vs current plugin/coding requirements).
- **Engineering velocity remains high** (month-to-date snapshot): **25 PRs opened / 18 merged**, **32 new issues**, **37 issues closed**, **27 active contributors**, **90 commits** (~**18.6k adds / 3.8k dels**, **160 files**). Core is advancing quickly, but UX/documentation and ops tooling are lagging.

---

## 1) Data Pattern Recognition

### Development velocity & trend
- **Throughput (monthly window 2026-02-01 → 2026-03-01)**:
  - PRs: **25 opened**, **18 merged** (72% merge rate)
  - Issues: **32 opened**, **37 closed** (net backlog down **-5**)
  - Contributors: **27 active**
- **Change profile**: heavy platform-level refactors + multi-language push (Rust/Python/TS) and runtime simplification (large v2 branches), alongside incremental hardening (null checks, auth, action filtering).
- **Risk signal**: multiple “large-scope” PRs/branches (e.g., v2.0.0 working branches) increase integration risk and can slow near-term UX fixes unless triaged into a stable release train.

### Community engagement patterns (Discord)
- **High-sensitivity topics dominate engagement**:
  - Token migration closure (deadline missed) generated repeated Q&A; migration/support channels now locked.
  - Token price speculation appears but is redirected to announcements (community seeking reassurance/clarity).
- **Support load concentrates on “ops tasks,” not dev questions**:
  - Cloud credit top-ups requested and solved manually (2 distinct requests in core-devs).
  - Account lookup friction for OAuth-created accounts (email search mismatch → needed org slug + Account ID).

### Feature adoption / usage indicators
- **Eliza Cloud**: users actively requesting credits implies usage; however, current process suggests *early-stage internal tooling* rather than self-serve billing/credits.
- **Automation ecosystem (n8n)**: active consolidation decision-making and build pipeline details indicate the plugin is moving toward being a “default” workflow automation path.
- **Social posting agents**: recurring integration pain (e.g., reported roomId/X posting issue) suggests adoption but brittle first-run UX.

### Pain point correlation across channels
- **Cost + duplication + reliability** form a cluster:
  - Open issue: URL in message triggers **duplicate LLM calls** (2x cost + duplicated output).
  - ActionFilterService introduced explicitly to reduce prompt bloat/costs.
- **Identity/account model complexity** shows up in both engineering and ops:
  - OAuth email lookup gaps (credits support).
  - Multi-entity/JWT auth work (recently implemented) indicates a broader multi-tenant direction—needs consistent identifiers surfaced to users/admins.

---

## 2) User Experience Intelligence

### Feedback themes (categorized by impact)

**A. Critical UX / Trust (High impact)**
- **Token migration missed-deadline experience**: user discovered too late; no remediation path; channels locked.
  - Sentiment risk: perceived “ruggy” even if process was fair—trust damage is driven by *communication + lack of graceful fallback*.

**B. “Getting started” mismatch (High impact)**
- Non-technical users asking for: zero-coding agent creation timeline, safe/recommended plugins, roadmap that reads without GitHub context, token utility/governance clarity.
  - Pattern: users compare against ChatGPT; current differentiation (“decentralized ownership + plugins”) is not landing without a guided path.

**C. Operational UX debt (Medium impact)**
- Credits administered by direct DB UPDATE; account discovery requires internal identifiers.
  - Pain: support bottleneck, risk of mistakes, lack of audit trail, inconsistent user/account resolution.

**D. Integration & reliability bugs (Medium → High impact depending on frequency)**
- Agent unable to post first message to X due to roomId-related error (unresolved).
- Cloud chat: image content stripped from LLM requests (open issue).
- Webapp: URL causes duplicate processing (open issue).

### Usage patterns vs intended design (observed)
- Users treat Discord as **primary support surface for billing/credits and token operations**, not only dev community. Current tooling doesn’t support that scale.
- Users expect **productized onboarding** (templates, one-click plugins, “what can I do today?” flows) more than framework-level primitives.

### Implementation opportunities (near-term)
- Convert repeated Discord explanations into **single canonical “ops FAQ”** (token migration status, credits, account identifiers).
- Make “internal-only” identifiers **visible in Cloud UI** (Account ID, org slug) to reduce support loops.

### Community sentiment (qualitative)
- Mixed: optimism on building and demos; frustration/confusion around token mechanics, deadlines, and “why ElizaOS vs ChatGPT” without plugins.

---

## 3) Strategic Prioritization (Impact × Risk × Dependencies)

### Priority 0 — Contain trust + support burden (1–3 days)
1) **Post-migration comms package (public, permanent)**
   - Deliverables:
     - A locked, linkable **Migration Postmortem/Status** page: deadline, 90-day window, ratio **1:6**, “no new tickets accepted,” what happens to pre-deadline tickets, and explicit “no exceptions” policy rationale.
     - Add **in-product banner** (Cloud + docs) for 14 days: “Migration closed; see status here.”
   - Impact: reduces repeated Discord load, prevents rumor spread, improves perceived fairness.
   - Risk: low technical risk; high comms value.

2) **Cloud credits: move from DB UPDATE to minimal admin tool + audit log**
   - Deliverables (MVP):
     - Admin-only endpoint/UI: “grant credits” with reason, actor, timestamp.
     - Audit table + export.
     - Account lookup by **email + OAuth subject + org slug + Account ID**.
   - Dependency: user identity model alignment (OAuth providers), DB schema.
   - Impact: removes “core-dev manual intervention” bottleneck; reduces errors/security risk.

### Priority 1 — Fix cost-amplifying UX bugs (this sprint)
Ranked by user pain × cost:
1) **Webapp URL double-processing → duplicate LLM calls** (open issue #6486)
   - Impact: **2x token cost** + visible duplicated output.
   - Recommendation: ensure URL is handled *either* as text *or* attachment before LLM invocation; add regression test with SSE stream assertions.

2) **Cloud chat image content stripped** (open issue #6494)
   - Impact: breaks multimodal promise; users perceive “it doesn’t work.”
   - Recommendation: patch `/api/v1/chat/completions` message conversion; add snapshot tests for mixed text+image messages.

3) **RoomId/X first-post failure**
   - Impact: kills “first success moment” for agent publishers.
   - Recommendation: add structured error reporting in UI + server logs; publish a troubleshooting doc; prioritize a fix if tied to recent runtime changes (“respond without roomId/worldId” work suggests shifting assumptions).

### Priority 2 — Make the product legible to non-coders (1–2 weeks, parallelizable)
1) **“Non-technical quickstart” funnel**
   - In Cloud dashboard: “Create your first useful agent” with 3 templates (e.g., personal inbox assistant, research summarizer, community moderator).
   - Include “recommended safe plugins” list and one-click enable where possible.
2) **Roadmap translation layer**
   - Convert GitHub roadmap to a **plain-language milestone board** (Now / Next / Later) including: zero-code timeline, token utility checkpoints, payment/billing milestones.
3) **Token utility/governance clarity**
   - Publish a concise token utility statement and what is *not decided yet*; reduce speculation vacuum.

### Priority 3 — Sustain dev velocity without destabilizing releases
- **Release train recommendation**:
  - Separate “v2 foundational refactor” from “UX/cost bugfix releases.”
  - Define a **stabilization branch** with strict scope: cost bugs, Cloud UX, auth/account tooling, critical plugin fixes.
- **n8n consolidation execution**
  - Complete repo consolidation around **plugin-n8n-workflow**; add migration notes; ensure build steps (“crawl” generated JSON) are impossible to miss (prepublish checks / CI guardrails).

---

## Resource Allocation Recommendation (next 2 weeks)
- **1 eng (platform/Cloud)**: credits admin tooling + account lookup normalization + audit logs.
- **1 eng (client/webapp)**: URL double-processing fix + SSE regression tests.
- **1 eng (API/runtime)**: cloud image message conversion fix + multimodal tests.
- **0.5 eng (devrel/docs)**: migration status page + non-technical quickstart + roadmap translation.
- **Optional (incentives)**: pilot “TipCat for explanations” on internal PR process to increase documentation quality without heavy process overhead.

---

## KPI Watchlist (start tracking weekly)
- **Support burden**
  - # of Discord threads about migration/credits per week
  - Median time-to-resolution for credit requests (target: <24h → then self-serve)
- **Cost/quality**
  - % of chats triggering duplicate LLM calls (target: 0)
  - Avg tokens per response before/after ActionFilterService rollout
- **Activation**
  - Cloud new accounts → first successful agent run rate
  - “Non-technical template” completion rate (if added)
- **Trust**
  - Migration-related sentiment mentions (qualitative) and recurrence after comms package is published