# ElizaOS Weekly Newsletter (2025-12-29) **Week of 2025-12-22 to 2025-12-28** ## 1) Executive Summary **1. Jeju + Babylon “decentralized Vercel” breakthrough** Core dev **Shaw** shared a major milestone: *cloud services running on both Jeju and Babylon*. The vision is a **compute marketplace** that automatically routes workloads to the *cheapest + fastest* resources available—aiming to deliver **~40% lower cloud bills** for web2 builders while enabling monetization via tokenization and network fees. Key pieces mentioned: agents + inference, DNS/domain purchasing, distributed cache, and a SQLite-compatible distributed DB layer (Upstash/Neon compatibility). **2. ElizaOS Cloud momentum + developer experience improvements** Community activity shows ElizaOS Cloud is being actively used to build and deploy agents, while ongoing work continues to smooth core UX (dashboard/agent builder) and improve real-time responsiveness (streaming/SSE). Even as token sentiment fluctuates, the technical roadmap is clearly accelerating toward “agents that can actually ship.” **3. Migration clarity (and pain points) surfaced and triaged** Discord saw continued friction around the **AI16Z → ElizaOS** migration (wallet connect issues, “max amount reached,” snapshot eligibility). The team and helpers consistently routed users into dedicated support channels while clarifying policy: **only balances held at the snapshot are eligible**—new purchases after the snapshot are not migrated. --- ## 2) Development Updates (Technical) ### Jeju network architecture: distributed hosting for “unstoppable agents” Shaw described Jeju as an ambitious distributed cloud platform designed to feel familiar to mainstream developers while being fundamentally decentralized: - Compatibility targets: **Vercel AI SDK**, **Kubernetes + Helm**, **S3**, **Redis/Upstash**, and **serverless SQLite** - Security posture: **Trusted Execution Environments (TEE)** with explicit attention to side-channel resistance - Key management: **distributed KMS** concepts (key sharding + proof-of-cloud), compared to approaches like Lit Protocol - Product direction: “**eliza cloud**” running on top of Jeju ### Real-time interaction: streaming + SSE (ecosystem trend) While GitHub activity in *Dec 28–29* was light, the broader weekly context remains clear: the ecosystem has been pushing toward **real-time, streamed agent responses** and more reliable message transport (including a shift toward **true SSE** instead of polling, per prior work tracked in the project’s weekly summary). ### Bugs and sharp edges reported this week Several actionable issues surfaced directly from builders: - **Agent naming bug:** names like `"null"` or numeric-only values (`"1"`, `"69"`, `"666"`) can fail to save or trigger client-side exceptions; `"null"` had odd behavior (editable in dashboard), while numeric names caused app errors. - **Streaming UI behavior:** streaming was added to a component, but multistep interaction UX still needs polish compared with “otaku” behavior. - **Deployment/auth friction:** users reported login and deployment errors (including a “username null” deployment issue), and discussion continued around implementing **custom SSO** to unify identity/auth across experiences. ### Database direction: “distributed SQLite” is becoming real Shaw is actively building a distributed SQLite layer and even asked for naming ideas (community suggestions included **“sq-lit”** and **“ShawQLite”**). This isn’t just a fun naming thread—it’s a signal that ElizaOS is working toward a developer-friendly default datastore that behaves like SQLite but scales across a distributed network. --- ## 3) Community Spotlight (Discord) ### Builders helping builders: migration support in action The most common support pattern this week was *fast triage → correct channel → reduce scam risk*: - **Omid Sa** directed users to the dedicated migration/support channel when wallet connection or eligibility questions came up. - **Borko** helped users who could see tokens but couldn’t exchange them. - **The Light** assisted users hitting the “max amount reached” wall. If you’re helping in Discord: keep doing what worked this week—**never accept DMs for support**, keep answers public, and route users to official channels. ### Ecosystem identity clarity: “by elizaOS” vs “built on elizaOS” Community members raised confusion around projects labeling themselves “by elizaOS.” A clear example: **Zora** was explicitly called **not officially affiliated** (unlike Spartan). This is an important ecosystem hygiene issue—clarity helps builders market themselves without accidentally implying endorsement. ### Partners channel: ambitious experiments In 🥇-partners, a developer teased a trading bot project (intentionally light on details), but it reflects what’s increasingly true: ElizaOS is attracting builders who want to deploy agents into real markets, not just demos. --- ## 4) Token Economics (AI16Z + auto.fun) ### Migration realities (and the “max amount reached” meaning) Discord repeatedly confirmed a key policy point: - **“Max amount reached”** commonly indicates the wallet **was not included in the snapshot**, or the user is attempting to migrate more than their snapshot-eligible amount. - **Post-snapshot purchases are not migrated** under current policy, which has been a major source of confusion for newcomers and exchange users. ### Sentiment snapshot: price down, distribution improving, utility narrative strengthening Community discussion acknowledged: - Token price is down **~90–98% from ATH**, causing frustration and hard questions. - Holder concentration improved: **top 10 holders decreased from ~89% to ~64%** (still concentrated, but trending healthier). - The narrative is shifting from “meme coin” to “utility platform,” with Jeju + Eliza Cloud as the credibility engine. ### auto.fun update No concrete auto.fun shipping notes landed in the provided activity this week. The practical takeaway: **watch for ecosystem monetization hooks to align with Jeju/Cloud (fees, deployment economics, marketplaces)**—that’s where auto.fun relevance is likely to become visible. --- ## 5) Coming Soon (What to Watch) - **Provider SLAs for the compute marketplace:** Shaw noted providers can offer SLAs even if the protocol doesn’t “guarantee” them itself—expect design work here. - **Fixes for migration UX and edge cases:** especially large-balance users seeing errors, and wallet connection limitations (e.g., Phantom visibility without exchangeability). - **Agent naming validation:** expect tighter server/client validation to prevent reserved words/numeric edge cases from breaking saves. - **File upload optimization for agents:** discussed as in-progress—important for richer agents with multi-file knowledge packs. - **Identity & SSO improvements:** ongoing discussion suggests auth consolidation is a priority as Cloud adoption grows. - **Distributed SQLite naming + integration:** once named, expect it to become a visible “default datastore” story for Jeju/Cloud apps. --- ## 6) Resources (Key Links) ### Discord threads / channels referenced - Core technical discussion (Jeju + Babylon cloud marketplace): https://discord.com/channels/1253563208833433701/1377726087789940836 - General discussion (migration questions + project affiliation clarity): https://discord.com/channels/1253563208833433701/1253563209462448241 - Partners channel (builder experiments): https://discord.com/channels/1253563208833433701/1301363808421543988 ### GitHub context to revisit - Weekly engineering summary (Dec 21–27): `github_summaries_week_latest_2025-12-21.md` (in your summaries feed) - Notable items mentioned in recent summaries: - SSE / real-time updates: https://github.com/elizaos/eliza/issues/5930 - Typed custom events (DX): https://github.com/elizaos/eliza/pull/6277 - Plugin Knowledge rate limiting: https://github.com/elizaos-plugins/plugin-knowledge/pull/49 - Snapshot eligibility / wallet support issue to track: https://github.com/elizaos/eliza/issues/6211 If you built something on ElizaOS Cloud this week, consider sharing it (and label it clearly as **“built on elizaOS Cloud”**)—it helps discovery while keeping ecosystem relationships honest.