# ElizaOS Weekly Newsletter (2026-03-13 → 2026-03-19) ## 1) Executive Summary **1. Two launches moved into the “imminent” window.** This week’s Discord updates pointed to the **Milady app** potentially launching immediately (discussed as “tonight”), with **Babylon** targeted for **early next week**. Community-created media (notably a deep-dive video) helped rally attention around Hyperscape, Babylon, and Milady as a cohesive push. **2. Agent persistence took a meaningful step forward with a new decentralized storage plugin.** A new community-announced plugin, **`@ensoul-network/plugin-elizaos`**, introduced an opinionated approach to “consciousness persistence” via encrypted storage, sharding, and validator-backed availability—plus a live explorer and quickstart docs. **3. Clearer signals on revenue and product strategy—amid heightened token frustration.** The team reiterated that **elizacloud is generating revenue** and is being used internally to build products. At the same time, token performance concerns escalated in chat, including renewed questions about buybacks and migration transparency. --- ## 2) Development Updates (Framework, Plugins, and Infra) ### New: Ensoul persistence plugin (`@ensoul-network/plugin-elizaos`) **DiamondRock – JD** shared a new persistence layer aimed at storing an agent’s identity + memory + learned behavior across a decentralized validator network. Key technical highlights discussed: - **Agent-owned encryption keys** and “**7 layers** of protection” (as described) - **Erasure coding + sharding** for resilience (reconstruct data if nodes fail) - **“Ensouled Handshake”** for cryptographic proof of persistent identity - A novel **“Consciousness Age”** metric (time continuously “ensouled” as a trust signal) - **Free storage for the first 100 agents** Why it matters: if you’re building agents that need continuity (identity, memory, trust) across deployments or restarts, this plugin proposes a drop-in path to persistence without central custody. ### Refactor: plugin naming standardization in xfn-framework To reduce confusion and make the registry easier to navigate, **Odilitime** noted a naming cleanup: - `plugin-form` → **`plugin-form-chain`** - `plugin-forms` (plural) → **`plugin-form`** (singular) - Registry references updated accordingly If you maintain integrations or docs referencing the old names, this is a good moment to update README snippets and install commands. ### Onchain data expansion: `plugin-evm` improvements **dinesh** reported work to: - Add **100+ onchain data support** via **goldrush.dev** integration - Fix open issues in **plugin-evm** This is a practical, developer-facing improvement: broader EVM data coverage tends to translate directly into better agent tooling for wallets, markets, contract introspection, and onchain analytics. ### v2.0.0 architecture discussion: skills without “plugin bloat” A key design conversation emerged around **ElizaOS v2.0.0** and PR **#6597** (skills folder structure). The underlying concern: the old 0.x era suffered from uncontrolled submissions and a noisy default surface area. A proposed direction: **ship v2.0.0 with zero default skills**, and instead support decentralized discovery—e.g., externally hosted **`skills.md`** files (example suggested: `yourdomains.com/skills.md`). This keeps core lean while still enabling an ecosystem of shareable skills. --- ## 3) Community Spotlight (Notable Contributions & Discussions) ### Seppmos’ deep-dive video: product narrative that connects the dots A community member shared a **deep-dive video** covering: - **Hyperscape** (agents battling onchain) - **Babylon** (agent prediction market game) - **Milady app** (positioned as an “agent hub”) This kind of “single story” content matters—especially when community sentiment is strained—because it translates disparate repos and prototypes into a coherent mental model for builders *and* token holders. ### “Unbrowse” integration proposal: a faster web for agents **lekt9** introduced **unbrowse**, a browser concept for agents that claims **~100x speed** by traversing APIs rather than DOM and indexing web APIs for agent use. **Odilitime** guided them to the plugin distribution path via the plugin registry and docs. If you’re building agent tools: the community is signaling strong interest in **agent-native browsing** that’s structured, high-throughput, and less brittle than UI automation. ### Directory/registry improvements (skills + plugins) **SYMBiEX (CidSociety)** proposed a more centralized directory (Clawhub-style) that houses **both skills and plugins** in a discoverable way. **Stan** voiced agreement—suggesting this is a ripe area for community-led organization. ### Support moments - **ElizaBAO** helped answer Milady launch timing questions from community members. - **Odilitime** repeatedly clarified revenue status, roadmap progress, and addressed communication pain points directly in chat. --- ## 4) Token Economics (AI16z, ElizaOS token, auto.fun) ### Market sentiment & buyback questions Discord saw sharp frustration about token performance (community members citing steep drawdowns and new lows). In response: - **elizacloud revenue** was confirmed as real and currently used for internal building. - **Buyback**: Odilitime indicated **buyback is part of the plan**, but **timeline/implementation is not yet confirmed** (“idk yet but that’s the plan”). Takeaway: buyback expectations exist, but there is no committed schedule in the captured discussions—plan accordingly. ### Migration transparency: unresolved questions on AI16z → elizaOS A serious thread raised concerns that a large portion of AI16z holders may not have migrated, with questions including: - What **percentage** actually migrated? - What is the **post-migration supply breakdown**? - What happens to **unmigrated allocation** originally intended for the community? - What will be done with **collected AI16z supply**? These were flagged as urgent by community members and remain **unanswered in the provided week’s logs**. ### auto.fun No concrete **auto.fun** development updates were captured in this week’s provided Discord summaries. If there are active threads elsewhere, consider linking them in next week’s edition so builders can track ecosystem tooling alongside token mechanics. --- ## 5) Coming Soon (What to Watch Next) - **Milady app**: discussed as potentially launching immediately; community estimates ranged from “tonight” to “next week or another week.” - **Babylon**: repeatedly stated target of **early next week**. - **Elizacloud go-to-market**: messaging suggests broader productization push by end of week / early next week (per earlier discussion). - **Skills distribution model decision**: expect follow-ups on whether v2.0.0 ships with **0 default skills** and how `skills.md` discovery should work. - **Operational fix**: investigate the **Google Form role assignment** issue (“Invalid Dynamic Link”) impacting onboarding. --- ## 6) Resources (Links & Pointers) - **Plugin registry (distribution + packaging):** https://github.com/elizaos-plugins/registry - **ElizaOS docs (plugin creation guide):** https://docs.elizaos.ai - **Ensoul persistence plugin docs (quickstart):** https://ensoul.dev/docs/quickstart.html - **Ensoul explorer:** https://explorer.ensoul.dev - **Ensoul repo:** https://github.com/suitandclaw/ensoul - **Discord discussion thread (weekly highlights source):** https://discord.com/channels/1253563208833433701/1253563209462448241