# ElizaOS Weekly Newsletter (Feb 1 – Feb 7, 2026)

## 1) Executive Summary

This week pushed ElizaOS forward on three fronts: **product launches**, **core reliability/security**, and **ecosystem expansion**.

- **Milaidy debuted as a new Mac-native Eliza-based assistant.** Built as an Eliza-flavored take on OpenClaw, *milaidy* aims to be a lightweight `.app` with a plugin-first architecture, agent “skills,” and OpenClaw connectors ported as Eliza plugins. The repo is live and already receiving community PRs and testing volunteers: https://github.com/milady-ai/milaidy

- **Babylon shipped to production and immediately hardened via rapid bugfixing.** The Babylon experience (agents + trading + feed) went production-ready, and the team merged a fix for a profile image upload bug within the same testing cycle—exactly the kind of tight feedback loop we want to see ahead of broader rollout.

- **Core runtime work accelerated toward “next-gen” Eliza.** Notable progress included: schema-driven, validation-aware prompt execution (merged), per-user request context for multi-tenant security (merged), plus continued momentum on a larger “next” branch spanning Rust/Python/TypeScript.

---

## 2) Development Updates (Core + Cloud + Apps)

### Core runtime & agent reliability

- **Dynamic execution engine (merged):** PR **#6384** introduced *schema-driven prompt execution* across TS/Python/Rust runtimes, including validation codes, retries with backoff, and streaming-aware extraction. The goal: make structured outputs more robust and detect partial/truncated generations before they silently break downstream logic.  
  Resource: https://github.com/elizaos/eliza/pull/6384

- **Per-user settings isolation via RequestContext (merged):** PR **#6457** added `RequestContext` (AsyncLocalStorage-backed) so `getSetting()` can resolve settings *per originating entity/user*, not just per agent. This is a major building block for **multi-tenant deployments**, preventing API key leakage and “shared agent” configuration collisions.  
  Resource: https://github.com/elizaos/eliza/pull/6457

- **Message bus event correctness (completed):** The long-standing issue where `EventType.MESSAGE_SENT` wasn’t emitted for messages submitted to the central server was fixed (PR **#6378**, related to issue **#5216**). This improves observability for plugins and downstream systems that depend on accurate lifecycle events.  
  Resource: (PR referenced in completed items) https://github.com/elizaos/eliza

- **Docs foundation strengthened (completed):** PR **#6356** added core documentation pages covering architecture, concepts, plugin dev, deployment, and API reference—important for onboarding the wave of new builders joining through Milaidy/Babylon momentum.

### “Next” / v2 direction

Two big branches remain in motion (not merged yet, but important signals):

- **“next” branch (PR #6474)** proposes a multi-language core (Rust, Python, TypeScript), examples for major frameworks, improved planning modes (including a “skip planning” fast path), and WIP “code agent” work.  
  Resource: https://github.com/elizaos/eliza/pull/6474

- **v2.0.0 working branch (PR #6351)** continues the direction of slimming down non-essentials (app/server/CLI) to focus on a portable runtime with critical plugins.

### ElizaCloud onboarding & account integrity (critical)

A high-impact onboarding issue was surfaced by **yojo**: welcome email links routed users to **dev.elizacloud.ai** instead of **elizacloud.ai**, leading to **duplicate accounts**, fragmented agents, and incorrect welcome credits (reported as $1 instead of $5). The team (Odilitime + sam) moved resolution to DMs for privacy and forwarded the bug to the cloud team.

Also reported: a **dashboard login loop** (cycling between login and dashboard), currently under investigation.

**Takeaway:** if you’re testing ElizaCloud onboarding, double-check which domain you’re on, and report anomalies early—these are conversion-killers at the exact moment users decide whether to pay.

---

## 3) Community Spotlight (Discord Highlights)

- **Wes jumped into Milaidy immediately with 3 PRs** and then paused to avoid overwhelming reviewers—requesting clarity on the preferred workflow (issues → PRs → tagging). This is a great example of “move fast, but coordinate.”

- **metalhorse233 shared a concrete Eliza-in-production use case:** they integrated Eliza into a game using **SpacetimeDB** as the backend and reported the setup was “surprisingly quick.” This is exactly the kind of implementation story we should document and replicate.

- **Babylon testing delivered fast results:**  
  - **ziflie** identified a profile image upload failure (“Failed to update profile”).  
  - **tcm390** implemented + merged a fix, and ziflie confirmed it working.  
  Tight loop, clean win.

- **kirsten launched BuildAMolt:** a hosted, private-VPS setup for Moltbots/Moltbook with ~2-minute provisioning, lowering friction for people who want reliability without local ops overhead.

- **Security concerns raised (needs follow-through):** DigitalDiva flagged possible malicious code in skills / setup vulnerabilities. No public technical write-up yet, but the signal is clear: as plugins/skills proliferate, we need stronger sandboxing, review paths, and provenance checks.

---

## 4) Token Economics (AI16Z / ELIZAOS + auto.fun)

### Migration & community expectations

Discussion remained active around the **recently closed 90-day migration window**. Community members reiterated that the timeline was communicated and extended, and that keeping migration tooling open indefinitely creates ongoing maintenance and support overhead. Several users asked what happens to unmigrated tokens; the answer shared: **locked for one year**.

### Market sentiment & marketing

Some holders expressed disappointment with price performance and called for a stronger marketing push ahead of upcoming launches. Others emphasized that stronger token performance benefits the ecosystem directly (treasury, runway, developer incentives), while also pointing to historical precedent: **ai16z previously ran from ~60k to 2.6b market cap in ~2–3 months** (shared as context, not a promise).

### auto.fun

No concrete, verifiable shipping update surfaced in the provided discussions this week. If auto.fun is part of your stack planning, keep an eye out for official announcements—and in the meantime, treat rumors as non-actionable until links and release notes drop.

---

## 5) Coming Soon (What to Watch Next)

- **Milaidy production push:** Mac `.app` packaging, skills parity with OpenClaw, and connector ports into Eliza plugins. Also: contribution workflow clarification so PR velocity doesn’t stall.

- **Cross-promotion strategy:** A live strategic question is how to market Milaidy without diluting the Eliza brand. Current direction: **separate brands, intentional cross-promotion**.

- **ElizaCloud fixes:** domain routing, account consolidation, correct credit assignment, and payment/recharge UX improvements (including requests for deposit addresses and fewer blockers like VPN payment failures).

- **Babylon expansion:** broader access beyond initial restrictions, continued reward/onboarding stabilization, and continued beta work on the Babylon “Monkey” game experience.

- **Security hardening:** stronger guardrails for untrusted skills/plugins, plus exploration of stateful sandbox infra (Sprites.dev was discussed as a model for hardware-isolated execution environments).

---

## 6) Resources (Links Worth Your Time)

- **Milaidy repo:** https://github.com/milady-ai/milaidy  
- **Dynamic execution engine (merged):** https://github.com/elizaos/eliza/pull/6384  
- **RequestContext / per-user settings (merged):** https://github.com/elizaos/eliza/pull/6457  
- **“next” multi-language core (in progress):** https://github.com/elizaos/eliza/pull/6474  
- **Discord threads referenced (high-signal):**  
  - coders channel: https://discord.com/channels/1253563208833433701/1300025221834739744  
  - core-devs channel: https://discord.com/channels/1253563208833433701/1377726087789940836  

If you shipped a fix, ran a real integration (like SpacetimeDB), or want to help test Milaidy/Babylon—post your findings with repro steps and screenshots. The community is moving fast right now, and clean reports are compounding progress.