## Intel — 2026-01-27 (covering activity through 2026-01-26)

### 1) Data Pattern Recognition

**Development velocity & trend**
- **Core throughput (month-to-date, 2026-01-01 → 2026-02-01 snapshot):** 37 PRs opened / **21 merged**, **88 issues opened / 50 closed**, **30 active contributors** (elizaos/eliza).
- **Recent daily execution (2026-01-26):** concentrated on **unblocking infra + enabling developer workflows**:
  - CLI import resolution fix (module resolution correctness)
  - AI Gateway credits blocking inference **resolved**
  - Core plugin integrations landed: **GitHub Plugin** + **Code Execution Plugin**
  - Extensive dependency patching on elizaos.github.io (maintenance-heavy day)

**Signals of execution risk**
- **Urgent ambiguity:** model naming convention confusion (Opus/Sonnet ↔ Ultra/Pro mapping) flagged as urgent; this is a *documentation + product correctness* risk that can cascade into support load and misconfigured deployments.
- **Behavioral mismatch reports:** plugin action handler callbacks behaving differently than docs expectation (only first callback visible; callbacks arriving at completion vs “immediate feedback”).

**Community engagement patterns**
- Discussion load skewed toward:
  - **Token price anxiety / migration questions** (high volume, recurring)
  - **Scam/security warnings** (ticket bot asking for wallets; “send tokens to migrate” scam pattern)
  - **Practical integration/how-to** (external MCP servers via Cloud API; self-hosting patterns on mini PCs / Oracle free tier)

**Feature adoption / usage proxies**
- Strong pull toward “**agent-as-a-service**” patterns:
  - External MCP server registration via `POST /api/v1/mcps` suggests developers want **tooling extensibility without forking core**.
  - Community self-hosting on low-power always-on machines suggests demand for **stable headless deployments** and **repeatable infra templates**.

**Pain point correlation across channels**
- **Docs correctness ↔ support load:** callback behavior + model naming confusion + token migration questions are all documentation-sensitive; each shows up as repeated Q/A rather than one-off bugs.
- **Security UX ↔ trust:** scam incidents and token anxiety amplify each other; lack of “single source of truth” messaging increases vulnerability to impersonation and social engineering.

---

### 2) User Experience Intelligence

**Feedback themes (categorized by impact)**

**A. High impact (blocks usage / trust)**
1) **Token migration clarity + scam resistance**
   - Users asking how to migrate from Ledger and from ai16z → ElizaOS; scammers actively intercepting with fake “support”.
   - Impact: potential direct user losses; reputational harm; support overhead.
2) **Model naming mapping ambiguity (Opus/Sonnet vs Ultra/Pro)**
   - Impact: incorrect model selection, cost/perf surprises, mis-benchmarked results, broken expectations in docs/tutorials.

**B. Medium impact (developer friction)**
3) **Plugin callbacks perceived broken / docs mismatch**
   - “Only first callback sent” + “callbacks arrive at completion” reports.
   - Likely interaction between **task planner mode (onestep vs multistep)** and UI/rendering of intermediate callbacks.
4) **Self-hosting patterns are emerging without official “gold path”**
   - Users independently using mini PCs / Oracle free tier. Lack of canonical guidance risks fragmented support and insecure configs.

**C. Opportunity signals (growth)**
5) **Platform distribution opportunity: Seeker phone DApp store**
   - Community sees current store quality as low; an “Eliza-as-Claude-style assistant” could become a standout app.
6) **Cloud extensibility demand: external MCP server support**
   - Clear developer appetite for registering external tools via API and letting Cloud proxy calls.

**Sentiment tracking (qualitative, with direction)**
- **Negative/Anxious:** token price ATL discussion; “lack of team response” perception.
- **Stabilizing via peers:** community members reassuring each other (4-year cycle framing).
- **Positive internally:** Sam’s return boosts execution confidence; Shaw rejoining Discord reduces “absent leadership” narrative (but needs reinforcement via proactive comms).

**Usage patterns vs intended design (notable mismatches)**
- Users expect **streaming/intermediate tool feedback** from action callbacks; current experience feels “batched at end,” reducing perceived responsiveness even if backend supports multiple callbacks.

---

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

#### Priority 0 (Trust & Safety) — do immediately
1) **Migration + anti-scam hardening**
   - **User impact:** Very high | **Technical risk:** Low–Medium | **Dependencies:** Comms + small product changes
   - Why now: active scams + repeated confusion = compounding losses.

**Recommended actions (48–72h)**
- Publish a pinned “**Official Migration & Support Policy**” message across Discord + docs:
  - “Support will never ask for your seed/wallet transfer.”
  - Only official domains + ticket workflow (explicit channel ID already used: `#1425417640071139358`).
- Add lightweight verification affordances:
  - A signed, versioned “official links” page; rotate and pin.
  - Consider Discord automod rules to block common scam phrases/addresses; add bot warnings when “wallet address” appears in support contexts.

#### Priority 1 (Product correctness) — unblock developer success
2) **Resolve model naming convention mapping (Opus/Sonnet ↔ Ultra/Pro)**
   - **User impact:** High | **Technical risk:** Low | **Dependency:** single decision + docs + config defaults
   - Output: one canonical mapping referenced by code, docs, UI labels, and CI workflows.

3) **Callback behavior: reproducible test + docs alignment**
   - **User impact:** Medium–High (dev UX) | **Technical risk:** Medium | **Dependencies:** task planner modes, message service, client rendering
   - Goal: ensure either (a) callbacks stream as documented, or (b) docs updated to reflect real semantics per planner/client.

**Recommended actions (this week)**
- Create a minimal repro matrix:
  - onestep vs multistep planner
  - SSE vs websocket vs discord transport
  - “callback count >1” actions
- Add an automated test asserting multi-callback ordering + timing expectations (or explicitly define “delivery timing contract”).

#### Priority 2 (Growth bets) — pursue with bounded scope
4) **External MCP servers + Cloud UX**
   - **User impact:** High for builders | **Technical risk:** Medium (security, tenancy) | **Dependencies:** auth, rate limits, secrets management roadmap
   - Tactical step: docs + example server + guardrails (timeouts, allowlists, audit logs).

5) **Seeker platform integration (distribution)**
   - **User impact:** Potentially high | **Technical risk:** Medium | **Dependencies:** app packaging, auth model, UX polish
   - Recommend a short discovery sprint rather than immediate build:
     - confirm Seeker APIs, app review, identity/sign-in, and whether “Claude-style assistant” UX is supported.

---

## Quantitative Summary (What to watch)
- **Dev capacity appears to be recovering**: key team member back (Sam) + infra blockers removed (AI Gateway credits).
- **Work mix risk**: high maintenance (dependency bumps) + multiple parallel roadmap issues can dilute focus; trust/safety items are currently generating the highest “cost of delay.”
- **Community risk indicators**:
  - Token anxiety + migration confusion + scams = elevated churn and reputational hazard even if core tech momentum is strong.

---

## Actionable Recommendations (next 7 days)

1) **Ship a “Trust Pack”**
   - One-page migration guide + anti-scam rules + official links registry.
   - Pin everywhere; update weekly until migration noise drops.
   - Success metric: migration questions in #discussion reduced; fewer ticket deflections; fewer scam reports.

2) **Close the model naming loop**
   - Decide canonical mapping; update:
     - docs, UI labels, environment variables, CI references, and any “Opus 4.5” workflow text.
   - Success metric: no new “is this right?” issues; reduced misconfiguration support.

3) **Callback contract: define, test, document**
   - If multiple callbacks are supported, ensure clients surface them incrementally (or document that they are “completion-phase” by design in onestep).
   - Success metric: Victor Creed’s repro resolved; docs no longer disputed; fewer repeated questions.

4) **Turn self-hosting into a supported path**
   - Publish an “official minimal self-host” guide aligned with observed behavior (mini PC + Oracle free tier profile).
   - Success metric: fewer fragmented setup questions; fewer inconsistent deployments.

5) **Bounded growth experiments**
   - Seeker: 1-week feasibility + prototype UX mock; no hard commitment until requirements clear.
   - MCP: add a reference “external MCP server” repo + security checklist (auth headers, request signing, rate limits, logging).

---

## Resource Allocation (pragmatic)
- **25–35%**: Trust & Safety + comms (migration/scams), until incident rate falls.
- **40–50%**: Developer experience correctness (model naming, callbacks, Cloud MCP docs/examples).
- **15–25%**: Growth experiments (Seeker feasibility, distribution packaging) with strict scope control to avoid roadmap dilution.