## Weekly Newsletter (2025-12-28 to 2026-01-03)

### 1) Executive Summary

This week was a solid “foundations + momentum” sprint for ElizaOS—tightening core reliability while the community pushed new ideas for agent UX and real-world integrations.

**Major milestones:**
- **MultiStep performance upgrade landed.** The default message service’s MultiStep provider execution was refactored to run providers in parallel (with fault tolerance), reducing latency for complex agents and workflows (PR **#6263**).
- **Critical SQL plugin fix to unblock data isolation.** A patch was delivered for PostgreSQL `SET LOCAL` commands by switching to `sql.raw()` to avoid parameterization errors that could break DB operations when `ENABLE_DATA_ISOLATION=true` (PR **#6316**).
- **Clear product/UX direction emerging for public agents + onboarding.** Several UI issues and proposals converged around improving the “public agent” experience (wallet connection flow, scrolling behavior, message limits for unauthenticated users, and clearer public-agent states).

---

### 2) Development Updates (Core + Plugins)

#### Core framework improvements (elizaos/eliza)
- **Faster MultiStep execution (PR #6263).** Providers in `runMultiStepCore` were moved from sequential execution to `Promise.allSettled`, meaning:
  - multiple providers can fetch in parallel,
  - failures don’t cascade (fault tolerance),
  - overall multi-step agent runs should feel snappier—especially when agents rely on multiple context sources/tools.
- **Logging consistency push (eliza/config).** In Discord, Stan shared that enhanced logging work is paired with a **log-format linter** in `eliza/config` to warn when logs don’t match expected structure. This is a developer-experience upgrade aimed at making debugging agents and plugins less chaotic as the ecosystem grows.
- **SQL plugin bug fix for Postgres `SET LOCAL` (PR #6316).** PostgreSQL `SET` statements don’t support parameterized queries; prior behavior could trigger errors like `syntax error at or near $1`. The fix uses inline values via `sql.raw()`, plus adds unit + integration tests to prevent regressions.

#### UX and product issues opened (and why they matter)
A cluster of GitHub issues opened by **@borisudovicic** signals a strong push toward a smoother “first contact” flow for new users:
- **#6317 – “Connect wallet should ideally go straight to wallet options.”** Less friction at the moment of conversion.
- **#6318 – “Scroll should work on whole page.”** Small detail, big impact on perceived polish.
- (Also visible in monthly roll-up) **#6312, #6313, #6314, #6315, #6311**—covering message limits for unauth’d users, separating public agent states, adding public agent metrics, adjusting free credits, and improving chat summaries.

#### Plugin ecosystem progress (context from weekly summary)
While the main aggregated activity here is core-repo heavy, last week’s cross-repo work sets up near-term improvements:
- **Twitter plugin:** OAuth2 PKCE work in progress for more secure auth.
- **OpenAI plugin:** media handling improvements + caching underway.
- **Registry:** continued expansion via community plugin submissions.

---

### 3) Community Spotlight (Discord Highlights)

#### “ElizaOS is the Linux of autonomous AI agents”
A recurring theme in discussion: ElizaOS is increasingly being explained as **an operating system-like framework** for agents—standardizing “infrastructure primitives” (DB, embeddings, model abstraction, APIs, tools) so builders can focus on the agent experience rather than plumbing.

This framing came up repeatedly in onboarding conversations, and it’s helping newcomers understand why ElizaOS is more than “just a wrapper” around model APIs.

#### roseOS: experimental agent framework built on ElizaOS
Community member **roseOS** introduced an experimental framework on top of ElizaOS focused on:
- **explicit agency boundaries,**
- **constraint-aware reasoning,**
- **accountability layers / traceability.**

This is exactly the kind of “advanced agent architecture” exploration that benefits from ElizaOS being composable—suggesting we’ll see more “agent frameworks built on the agent framework.”

#### Multi-model agents: practical guidance from core devs
In `#coders`, **Stan** helped clarify how to run **multiple models in a single agent** (e.g., Anthropic for forecasting/calculation-style tasks, OpenAI for general reasoning):
- Recommended approach: use the **OpenRouter plugin** and define provider/model routing in environment configuration.
- Deployment note: for agents with custom plugins, **cloud containers** were recommended as the deployment path.

#### Gaming + worldbuilding thread continues
On Jan 1, the community continued debating the “Zelda with agents” direction—forking a RuneScape project and using Eliza agents to populate towns, generate quests, and drive NPC behavior. The discussion wasn’t just hype: it produced a concrete list of needs (quest generation, NPC setup, world viewing/tweaking) that map well to future plugin + tool development.

---

### 4) Token Economics (AI16z token + auto.fun)

This week included **community questions** about token migration and ecosystem mechanics, but **no official, finalized announcement** landed in the provided sources.

What we *did* see:
- Continued user uncertainty around **migration from AI16Z to ElizaOS** (questions raised in Discord, largely unanswered in-thread).
- Product/monetization adjacent discussion surfaced on GitHub:
  - **#6315 – “Change free credits from $5 to $1.”** This is not a token change, but it *does* reflect ongoing tuning of onboarding economics and cost control.
- **auto.fun:** no explicit weekly development notes were included in the aggregated data for this period. If you’re tracking auto.fun specifically, keep an eye on official announcements and pinned Discord updates as they appear.

**Community takeaway:** if you’re planning anything that depends on migration mechanics (eligibility, wallet flows, snapshots), treat current chatter as exploratory until it’s confirmed via official docs or announcements.

---

### 5) Coming Soon (What to Anticipate)

Based on open issues + active discussion, expect movement on:

- **Public agent UX redesign:** clearer separation of states (unauth visitor vs signed-in non-owner vs owner), better share/fork flows, and likely analytics primitives (views/chats/forks).
- **Onboarding gating and incentives:** message limits for unauthenticated users, refined free credits, and a smoother wallet connection path.
- **More robust developer observability:** wider adoption of the logging linter + standardized log formats across plugins.
- **Multi-model patterns becoming “official”:** community interest is high—examples/docs for multi-model agent routing would help standardize best practice.

---

### 6) Resources (Links)

**Roadmap**
- ElizaOS roadmap: https://github.com/elizaOS/roadmap

**Key Pull Requests**
- PR #6263 – MultiStep provider parallelization: https://github.com/elizaOS/eliza/pull/6263  
- PR #6316 – SQL plugin fix for `SET LOCAL` using `sql.raw()`: https://github.com/elizaOS/eliza/pull/6316  
- PR #6301 – License year update (maintenance): https://github.com/elizaOS/eliza/pull/6301

**Key Issues (UX + onboarding)**
- #6317 – Wallet connect should go straight to options: https://github.com/elizaOS/eliza/issues/6317  
- #6318 – Scroll should work on whole page: https://github.com/elizaOS/eliza/issues/6318

**Discord discussion hubs referenced**
- General discussion: https://discord.com/channels/1253563208833433701/1253563209462448241  
- Coders channel: https://discord.com/channels/1253563208833433701/1300025221834739744