## User Feedback Analysis — 2026-02-10 (based on 2026-02-07 to 2026-02-09 Discord + recent GitHub issues) ### Data Snapshot (for quantification context) - **Primary sources reviewed:** Discord channels (**💬-discussion**, **💬-coders**, **core-devs**, **🥇-partners**) across Feb 7–9; GitHub issue/PR highlights from the Feb monthly rollup. - **Observed mix:** Community sentiment (token/project direction), install/runtime bugs (milaidy + plugins), performance/infrastructure concerns, and product/UX requests (ElizaCloud + character tooling). --- ## 1) Pain Point Categorization (Top recurring friction areas) ### 1. Community — Token anxiety + trust/communication breakdown (highest severity, highest frequency) **What users reported** - Ongoing **token price decline** and “no utility” framing dominated **💬-discussion** across multiple threads. - **Rumor-driven panic**: confusion about **Korean exchange delistings** (Bithumb/Coinone/Korbit) until clarified that **AI16Z (pre-rebrand)** was delisted, not **ELIZAOS** (clarified by **davidhq**, **paolin**). - **Team presence expectations**: repeated complaints that “admins aren’t here,” “devs left,” “what happened to promised products,” plus allegations of insider selling (wallet-tracking narrative around an address alleged to be Shaw’s; disputed by **ovo**, **davidhq** requesting proof). - A **critical token migration question** (“missed ai16z→ELIZA migration deadline—are tokens lost?”) was asked by **ufw** and **went unanswered**, amplifying distrust. **Who it impacts most** - Token holders and newcomers deciding whether to commit time/money; also builders affected indirectly because the conversation crowds out technical support. --- ### 2. Technical Functionality — Fresh install crashes & runtime robustness gaps (high severity, high frequency in coders) **What users reported** - **Crash on fresh milaidy VPS install**: `skill.description.toLowerCase is not a function` in `agent_skill_instructions` provider (reported by **azsxdc**, confirmed known bug by **Odilitime**). - Multiple confusing errors on new installs: - “server id missing” / “ownership” errors (stated as **normal/ignorable** by **Odilitime**) - `pluginRegistryService` **401 Unauthorized** (called **WIP functionality**) - GitHub corroboration of robustness issues: multiple recent fixes around **null/undefined checks** in bootstrap/providers/settings (PRs like defensive checks), indicating this is a recurring class of runtime failures. **Who it impacts most** - New developers trying milaidy/openclaw setups, especially on VPS; anyone evaluating ElizaOS quickly (“time-to-first-success” is low). --- ### 3. Integration — Plugin compatibility, wallet tooling gaps, and ecosystem uncertainty (high frequency, medium-to-high severity) **What users reported** - **Plugin compatibility issues** with milaidy and missing constants (e.g., **MAX_EMBEDDING_TOKENS** discussed Feb 8). - Wallet-related integration questions: - “best wallet management plugin” for openclaw game skills (asked by **kira**, unanswered) - “Cardano wallet integration—do we need it?” (asked by **Wes**, unanswered) - “bridge ELIZA Solana→BSC?” answered clearly: **no official/safe bridge** (by **EMERSON3S**)—users are actively trying cross-chain moves. - Builders promoting/paywalling agent experiences (e.g., hackathon project using **x402/SOL paywall** + Polymarket scanning) suggests payments/wallets are a real usage axis, but guidance/tooling is thin. **Who it impacts most** - Builders shipping agent-native games and monetized agents; users trying to move tokens/assets across chains; plugin authors needing stable contracts. --- ### 4. Performance — Database/logs bottleneck slowing responses (high severity, medium frequency) **What users reported** - Postgres **logs table hit excessively → slower Eliza response times** (Feb 8); **Stan** confirmed a “major optimization plan” is in progress. **Who it impacts most** - Cloud/self-host users at any scale; this becomes a platform-level reliability perception issue (slow agents feel “broken”). --- ### 5. UX/UI — Character configuration & model selection mismatch (medium severity, medium frequency) **What users reported** - Agents **defaulting to “eliza” personality** instead of custom characters; limited creation path (“only via onboarding”) until a character editor is wired in (per **Bill Ding**). - Model selection mismatch reported (e.g., Claude Haiku used despite config) on Feb 8. - ElizaCloud feedback: requests for **prompt-only (zero coding) UX**, better visual generation flow, and a plugin marketplace + curated plugin bundles (from **yojo**). **Who it impacts most** - Non-technical creators and product users; also developers trying to demo agents with reliable “persona control.” --- ### 6. Documentation — Missing “official answers” to high-stakes questions (medium severity, high frequency) **What users reported** - Repeated requests for: roadmap, updates, token utility, staking, migration, delisting clarification. - “Where can we read updates?” answered via **weekly roundup videos**, but users still express they can’t find canonical written guidance quickly. - Confusion over which errors are normal vs critical on installs indicates missing “first-run known warnings” docs. **Who it impacts most** - Newcomers, token holders, and builders troubleshooting setups. --- ### 7. Security — Scams and unsafe support surfaces (high severity, medium frequency) **What users reported** - Scam attempts and a fraudulent “support desk discord” (confirmed scam by **Monsgroow.**). - Call for better **autoban systems** for scam tickets (raised by **azsxdc**, acknowledged by **Odilitime**). **Who it impacts most** - New users (highest risk) and community moderators (high load). --- ## 2) Usage Pattern Analysis (actual vs intended) ### Observed real-world usage 1. **Agent-native games & automated “operator” bots** - Hackathon project: Polymarket signal scanning + autoposters + paywalls. - SpacetimeDB game integration reported as “surprisingly quick” (Feb 7). 2. **“Personal assistant as an app” direction** - **milaidy** positioned as a Mac-native minimal-bloat app with plugin-based connectors—users are rallying around “simple install + daily use.” 3. **Social integrations as primary value** - Twitter posting works but quote-repost limitations; Discord/Telegram/SMS messaging integrations show up in broader project direction. 4. **Non-technical / prompt-only creator demand (ElizaCloud)** - Clear request trend: “let me do value without coding,” plus curated plugin experiences. ### Mismatches vs intended/assumed usage - The framework is evolving quickly (multi-language v2 work, plugin bootstrap changes), but many users approach it as a **product** (ElizaCloud / milaity app) and expect **stable UX + clear official policies** (bridges, migration, staking, exchange status). - Users interpret dev velocity through token price; they expect **token-holder comms** akin to a crypto product, not an open-source framework. ### Feature requests aligning with actual usage - **Plugin marketplace + curated bundles** (matches non-technical prompt-only demand). - **Wallet/payments primitives** (matches paywalled agents + game economies). - **Character editor + persona reliability** (matches creators shipping agent “characters”). - **SillyTavern character card support** (matches roleplay/character-driven agent creation). --- ## 3) Implementation Opportunities (solutions per major pain point) ### A) Token/community trust & communication (Community) **Opportunity 1 (High impact / Low difficulty): Ship a single “canonical status hub”** - Create a pinned, always-updated **“What’s True Today”** post covering: - exchange listing/delisting status (AI16Z vs ELIZAOS) - bridges (official vs unsafe) - staking (available/not) - migration (open/closed + remediation path) - Model: many projects use a **Statuspage + pinned FAQ** (e.g., incident-style updates even for non-incidents). **Opportunity 2 (High impact / Medium difficulty): Weekly “build + business” changelog in text** - Convert the “Cron Job” video into a short Markdown digest with: - shipped items, WIP, next week goals, and “known issues” - explicit answers to top recurring questions - Model: Rust, Next.js, and many OSS projects ship **release notes + “known issues”** sections. **Opportunity 3 (Medium impact / Medium difficulty): Formalize token-holder Q&A cadence** - Monthly AMA with pre-collected questions; publish transcript and decisions. - Add “Decision log” to reduce rumor cycles (“why no staking yet,” “what token utility is in-scope”). --- ### B) Fresh install crashes & confusing startup errors (Technical Functionality + Documentation) **Opportunity 1 (High impact / Low-to-medium difficulty): Harden skill providers & validate schemas** - Fix `skill.description` type assumptions: ensure string coercion or schema validation at load time. - Add a validation step that outputs actionable error messages (“Skill description must be string; got object in X skill file”). **Opportunity 2 (High impact / Low difficulty): “First-run warnings” guide + log classifier** - Document which errors are safe to ignore (server id/ownership), which are WIP (401 registry), which are fatal. - Implement a simple **log classifier** that prints: `KNOWN_WARNING`, `ACTION_REQUIRED`, `BUG` with links to docs/issues. - Model: Terraform and Kubernetes CLIs often map common failures to docs URLs. **Opportunity 3 (Medium impact / Medium difficulty): Provide a single “golden path” deployment** - One supported install target at a time (e.g., “milaidy on VPS” or “local dev”), with a tested script/compose. - Reduce combinatorial setup complexity; publish “known-good versions” matrix (core + plugin versions). --- ### C) Plugin ecosystem friction & wallet/payment gaps (Integration) **Opportunity 1 (High impact / Medium difficulty): Publish a stable plugin compatibility contract** - Versioned API contract for plugins (capabilities, required constants like MAX_EMBEDDING_TOKENS). - Add CI that checks required exports across plugin repos. **Opportunity 2 (High impact / Medium difficulty): Create “Wallet & Payments” reference implementation** - A minimal, audited wallet management plugin pattern: - key storage guidance (per-user connections / data isolation) - transaction guardrails (leverage existing validation/guardrail direction) - Model: many ecosystems provide a “reference adapter” (e.g., OAuth reference providers, Stripe sample apps). **Opportunity 3 (Medium impact / Low difficulty): Answer unanswered wallet questions with a decision** - Even “Not supported / out of scope for now” is valuable: - Cardano: supported now/next/later? - Openclaw wallet mgmt plugin: recommended approach? - Convert those Qs into GitHub issues with labels + owners. --- ### D) Database/log performance bottleneck (Performance) **Opportunity 1 (High impact / Medium difficulty): Reduce write amplification + add log retention** - Implement log sampling or async batching; add TTL/partitioning for logs table. - Default retention policy for self-host and cloud. **Opportunity 2 (High impact / Medium difficulty): Add performance budgets & tracing** - Add tracing around message handling and DB queries; expose “slow query” diagnostics. - Model: Postgres-heavy apps often ship with **pg_stat_statements + dashboards** and app-level spans. **Opportunity 3 (Medium impact / Low difficulty): Document scaling guidance** - “If responses slow down, check X; recommended indexes; recommended retention.” --- ### E) Character/persona reliability and ElizaCloud prompt-only UX (UX/UI) **Opportunity 1 (High impact / Medium difficulty): Ship the character editor + import/export** - Include import support for formats like **SillyTavern cards** (already being built by Vega) to reduce friction. - Provide deterministic preview: “This is the active system/persona + model.” **Opportunity 2 (High impact / Medium difficulty): Make model selection observable** - Display active model in UI and logs; warn when fallback occurs. **Opportunity 3 (Medium impact / Medium difficulty): Curated plugin marketplace / “recipes”** - Provide “certified plug-and-play” bundles and upsell packages (as requested), plus community ratings. - Model: VS Code Marketplace + “recommended extensions,” or Home Assistant’s curated integrations. --- ### F) Scam resistance (Security) **Opportunity 1 (High impact / Low difficulty): Pinned security banner + official support verification** - “Official support will never DM first,” list official links, and a “verify support server” checklist. **Opportunity 2 (High impact / Medium difficulty): Automod/autoban for scam patterns** - Add regex + reputation gating for new accounts posting “support desk” links; quarantine suspicious posts. **Opportunity 3 (Medium impact / Medium difficulty): Incident workflow** - A simple mod playbook (“when scam detected, do X; post notice; remove links; ban”). --- ## 4) Communication Gaps (expectations vs reality) ### Repeated expectation mismatches - **“Is this a product with support?”** Users act as if there is guaranteed support and official bridges/staking; reality is partial/none (“no official bridge,” “no staking”). - **“If errors appear on first run, am I broken?”** Normal warnings are not documented; users assume failure. - **“Token migration deadline—what now?”** High-stakes user question went unanswered; expectation is that deadlines have remediation. ### Recurring questions indicating onboarding/documentation gaps - Roadmap visibility (“can I see roadmap?” unanswered) - Migration status/remediation (unanswered) - Wallet management guidance (unanswered) - Delisting clarification needed official statement despite community clarification - Where to get updates (videos exist, but users want a canonical written hub) ### Specific documentation improvements - Add a **“Critical Token/Exchange FAQ”** page (delisting, migration, bridges, staking). - Add **“First Install: Known Warnings vs Real Errors”** page with copy-pastable log snippets. - Add a **“Wallet & Payments: Supported Patterns”** guide (even if minimal). --- ## 5) Community Engagement Insights ### Power users (and their needs) - **Odilitime**: frontline support + troubleshooting; needs better tooling/docs to reduce repeated explanations (log classifier, known-issues doc). - **Stan**: performance/infrastructure direction; needs a way to broadcast progress in a format users will read (text digest + status hub). - **Bill Ding**: product UX answers (character editor, Twitter plugin); needs issue intake that turns “no idea, file an issue” into structured tasks. - **jin**: publishing updates; needs a dedicated updates channel + transcript workflow so updates are searchable. - **davidhq / paolin / ovo**: rumor control and trust defense; need an official comms artifact to point to, rather than re-litigating in chat. ### Newcomer friction signals - Fresh VPS install failures (azsxdc) + confusion about which errors matter. - Scam exposure (ufw) while looking for support. - Token migration missed deadline question (ufw) showing that key lifecycle events aren’t reaching users reliably. ### Converting passive users into contributors - Create a **“good first issue: docs”** track: - “First-run warnings doc” - “Token FAQ page” - “Install guide for VPS” - Create a **maintainer triage hour** weekly: move unanswered Discord questions into GitHub issues with labels/owners. - Recognize and empower volunteer testers (milaidy volunteers) with a clear contribution workflow (Wes explicitly paused PRs pending clarity). --- ## 6) Feedback Collection Improvements ### Effectiveness of current channels - **Discord** is effective for rapid discovery but poor for resolution: - multiple **unanswered high-stakes questions** (migration, wallet mgmt, roadmap) - repeated rumor cycles (delisting) due to lack of a canonical reference - **GitHub** captures technical issues well (e.g., URL triggers duplicate LLM calls), but many user questions never become issues. ### Improvements for structured, actionable feedback - Add a **Discord bot command** to file issues: `/issue create` pre-fills template (error logs, environment, expected vs actual). - Weekly “Top 10 unanswered questions” digest posted by a bot to prompt maintainers/mods. - Add short **in-product feedback prompts** in ElizaCloud (“What blocked you today?” with categories). ### Underrepresented segments - **Non-technical ElizaCloud users**: only a small amount of direct feedback (notably yojo), but requests imply a larger silent cohort. - **VPS/self-host operators**: we see pain via azsxdc, but need more systematic telemetry/feedback (install success rate, common failure points). --- ## Prioritized High-Impact Actions (next 2–4 weeks) 1) **Publish a canonical “Status & FAQ” hub** (migration, delisting clarification, bridges, staking, official links) and pin it across Discord channels. 2) **Fix the milaidy agent skill provider crash** (`skill.description.toLowerCase`) and add schema validation + a “known warnings vs critical errors” install guide. 3) **Implement basic anti-scam defenses**: pinned security banner + automod rules for common scam patterns; add an official support verification flow. 4) **Ship a text-based weekly changelog** (in addition to video) with “shipped / WIP / next / known issues,” linked from the FAQ hub. 5) **Define and publish a plugin compatibility contract** (including required constants) and add CI checks to prevent ecosystem drift; start with the most common milaidy/openclaw plugin paths.