# ElizaOS Intel — 2026-04-13

## 0) Exec Snapshot (last observed + rolling MTD)
**Development throughput (MTD 2026-04-01 → 2026-05-01 window, as of today’s dataset):**
- **PRs:** 4 opened, **3 merged**
- **Issues:** 7 opened, **1 closed**
- **Contributors:** **14 active**
- **Code churn:** **+6,785 / −3,732** across **81 files**, **70 commits**  
**Signal:** High churn concentrated in core runtime + dev harness changes; elevated regression risk and documentation drift.

**Community/Support signal (latest Discord data: 2026-04-10):**
- High-intent developer integration work ongoing (Socket.IO protocol reverse-engineering).
- Token-economics questions generating **frustration due to unclear timelines** (airdrop/buyback/Jeju gas fees).
- Wallet verification support issue (Collab.land) **unresolved in-thread**.

---

## 1) Data Pattern Recognition

### 1.1 Development velocity & trend
- **Core runtime changes landed quickly** (notably PR **#6562 merged 2026-04-08**, PR **#6702 merged 2026-04-09**), followed by additional pending core behavior changes (PR **#6712** addressee routing; PR **#6709** TOON/action params fix).
- Trend is **“ship core refactors fast, stabilize later”**: large deltas, multiple behavioral surfaces (messaging pipeline, shouldRespond, logging, memory persistence controls, prompt templates) evolving simultaneously.
- **Risk profile increasing** due to:
  - Merged changes with **known reviewer-flagged logic issues** (see Strategic Prioritization: memory IGNORE persistence + embedding zero-vector).
  - Pending PRs that alter routing/speaking behavior in group contexts (PR #6712).

### 1.2 Community engagement patterns
- Developer engagement is **hands-on + integration-driven** (Socket.IO dashboards, local model hosting, fallback transports).
- Product/community engagement is **economics-driven** (utility/buyback/gas fees) with **sentiment degradation** when answers are non-committal.
- Support engagement shows **leakage**: wallet verification issue did not reach resolution path or documented runbook.

### 1.3 Feature adoption & usage vs. intended design
- **TOON encapsulation + connector compatibility** is actively used (Discord connectors / “milady” connector), and missing schema fields caused real failures until PR #6709.
- Users are still investing in **v2 Socket.IO** despite team messaging that focus is on **v3**, indicating a mismatch between roadmap focus and ecosystem dependency reality (integrators need stable protocols even if not strategic).

### 1.4 Pain point correlation across channels
| Pain point | Where it shows up | Correlated impact |
|---|---|---|
| Missing/implicit protocol documentation (Socket.IO message types/fields, DM creation) | Discord dev support | Slower integrations, repeated Q&A, increased support load |
| Token roadmap ambiguity (airdrop/buyback/Jeju fees) | Discord community | Sentiment decline; reputational risk; reduces willingness to build |
| Onboarding/env breakages (macOS bun postinstall; Windows checkout artifacts; submodule/workspace complexity) | GitHub issues/PR review notes | Direct adoption friction; increases time-to-first-agent |
| Identity/trust direction shift (AgentID, capability-based auth) | Weekly summary + open proposals | Requires sequencing; breaks expectations of “messaging-first” users |

---

## 2) User Experience Intelligence

### 2.1 Feedback categorization by impact & theme

**A) Critical UX friction (blocks adoption)**
1) **`elizaos create` fails on macOS due to bun dependency/postinstall behavior** (Issue **#6704**, open)  
   - Outcome: CLI cleans up and leaves nothing → “hard fail” with no artifact to debug.
2) **Dev harness/submodule workflow fragility** (PR **#6702** merged but review flagged workspace/lock inconsistencies)  
   - Outcome: fresh clones/CI/dev machines may fail installs or resolve wrong deps.

**B) High-impact behavior correctness (user-visible agent behavior)**
1) **Group-room routing / addressee resolution** (PR **#6712**, pending)  
   - Review flagged **P1 bug**: aliasEntity ambiguity breaks address resolution when entityId ≠ agentId, undermining the feature.
2) **Async action continuation loop spam** fixed (PR **#6709**, pending)  
   - Strong UX win: reduces filler/duplicate responses while tasks run.

**C) Trust & governance (credibility / long-horizon adoption)**
- **Token-econ clarity**: repeated questions with “no new details” response pattern → community frustration.
- **Wallet verification support**: unresolved Collab.land issue suggests missing escalation path/runbook.

### 2.2 Implementation opportunities (low/medium lift)
- **Codify Socket.IO v2 protocol** minimally: enums + required/optional fields + example flows (join room, send message, auth, create DM).
- **Support runbooks**: Collab.land verification troubleshooting checklist + escalation route.
- **CLI failure ergonomics**: preserve generated project folder on failure, emit “next steps” with deterministic remediation.

### 2.3 Community sentiment (qualitative)
- Dev sentiment: productive but blocked by missing docs; workaround culture (“use Cursor to query code”) indicates docs lag.
- Investor/user sentiment: frustrated by timelines; needs proactive comms with concrete milestones or explicitly scoped non-commitments.

---

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

### 3.1 Highest priority: stabilize merged core changes with known correctness risks
**Why now:** Core runtime reliability is the multiplier for every plugin, connector, and v3 migration path.

1) **Memory persistence correctness regressions (merged PR #6562)**
- Reviewer-flagged:
  - **IGNORE-path memory persistence bug** when `ALLOW_MEMORY_SOURCE_IDS` is set (hardcoded `"agent_response"` check) → silent data loss for IGNORE decisions.
  - **Embedding failure stores zero-vector** → “stored but unretrievable” semantic memory corruption.
- **User impact:** High (memory quality + debugging trust).  
- **Technical risk:** High (data integrity).  
**Action:** Ship a hotfix PR with:
  - Correct allowlist logic for IGNORE branch (match non-IGNORE behavior).
  - On embedding failure: do not persist vectorized memory as zero-vector; instead mark as “unembedded” and retry/backfill or store in separate queue.

### 3.2 Next priority: unblock adoption by fixing onboarding/platform blockers
2) **macOS `elizaos create` bun dependency issue (Issue #6704)**
- **User impact:** Very high (blocks first-time users).
- **Fix scope:** Medium (dependency graph + packaging policy).  
**Action:** Remove `bun` as runtime dependency in CLI/bootstrap where possible; add detection + message:
  - Detect bun “postinstall not run” scenario.
  - Offer automated remediation: rerun install script or switch to npm/pnpm instructions.

3) **Dev harness/submodule workflow hardening (post-merge follow-up to PR #6702)**
- **User impact:** High for contributors; moderate for end users.
- **Risk:** Medium-high due to workspace resolution variance.  
**Action:** Add CI job that runs on a **fresh clone without submodules initialized** and validates:
  - `bun install`
  - `bun run dev` (or a dry-run)
  - ensures `package.json` workspace entries are consistent with committed state.

### 3.3 Protect UX: ensure pending behavior-changing PRs are correct before merge
4) **PR #6712 group addressee routing**
- **User impact:** High (when agents speak/ignore in multi-party rooms).
- **Risk:** Medium-high (core shouldRespond path + prompt regen across TS/Python/Rust).  
**Dependency:** Fix aliasEntity ambiguity; add missing test coverage for entityId ≠ agentId case.  
**Action:** Block merge until:
  - aliasEntity fix landed
  - regression tests added for mixed IDs and multiple agents with overlapping name tokens.

5) **PR #6709 TOON params + async terminal action continuation**
- **User impact:** High for connector integrations; reduces spam.
- **Risk:** Low-medium (broad template migration noted, but fixes are additive).  
**Action:** Prioritize review/merge after #6712 is stabilized (to avoid stacking behavior changes in message pipeline simultaneously).

### 3.4 Community trust: reduce token-econ ambiguity drag
6) **Token comms package (airdrop/buyback/Jeju fees)**
- **User impact:** High (sentiment + retention).
- **Risk:** Low technically, medium reputational if overpromised.  
**Action:** Publish a single pinned “Token Status” post with:
  - What is confirmed vs. not decided
  - Explicit dependencies (e.g., v3 milestones, treasury policy, legal constraints)
  - Next update date (even if “no change,” it reduces repetitive threads)

---

## 4) Critical Path Dependencies & Resource Allocation

### Critical path
1) **Runtime correctness + memory integrity** (hotfix #6562 fallout) → prerequisite for AgentID/trust layer credibility.
2) **Onboarding stability (CLI + dev workflows)** → prerequisite for ecosystem growth (plugins, connectors).
3) **Group routing correctness (#6712)** → prerequisite for multi-agent rooms, reduced loop/spam.
4) **Protocol documentation (v2 Socket.IO)**
   - Even if v3 is priority, integrators are building now; lack of docs becomes ongoing support tax.

### Recommended allocation (next 1–2 weeks)
- **40% Core runtime stabilization:** memory integrity hotfixes, regression tests, P99 latency guardrails.
- **30% DX/onboarding:** CLI macOS fix + fresh-clone CI + submodule/workspace documentation.
- **20% Multi-party behavior correctness:** #6712 bugfix + tests; merge sequencing to avoid stacked regressions.
- **10% Comms/support ops:** token status pin + Collab.land runbook + escalation path.

---

## 5) Actionable Recommendations (tracked outcomes)
1) **Ship “Memory Integrity Patch”** (target: 48–72h)
   - Fix IGNORE allowlist logic + embedding failure handling
   - Add tests: ALLOW_MEMORY_SOURCE_IDS + IGNORE path; embedding failure does not corrupt recall
2) **Close macOS CLI blocker** (target: 3–5 days)
   - Remove bun runtime dependency; improve failure UX (don’t delete project; print deterministic fix steps)
3) **Gate PR #6712 on correctness** (target: before merge)
   - Fix aliasEntity; add entityId/agentId mismatch tests
4) **Publish v2 Socket.IO “minimal canonical spec”** (target: 1 week)
   - Message types, auth requirements (entityId UUID), join/send examples, DM creation guidance (or explicit “not supported”)
5) **Reduce support leakage**
   - Add Collab.land verification troubleshooting doc + named maintainer rotation for unresolved support threads
6) **Token comms**
   - Pinned status + next-update timestamp to reduce repetitive Q&A and sentiment decay