# elizaOS Discord - 2025-03-22

## Overall Discussion Highlights

### ElizaOS Development Updates
- **V1 vs V2 Differences**: Jin shared a blog post explaining differences between ElizaOS v1 and v2. The underlying LLM model remains the same between versions, but the way it "experiences the world" is changing.
- **Demo Day Announcement**: Jin announced an upcoming demo day for ElizaOS.
- **V2 Launch Timeline**: The community expects v2 launch, launchpad implementation, and marketing efforts within approximately two weeks.

### Technical Issues & Development
- **Plugin Problems**: Multiple users reported issues with plugin loading failures, particularly with plugin-sql in beta versions (1.0.0-beta.7).
- **Agent Personality Persistence**: Several users reported issues with agent personalities persisting incorrectly despite configuration changes, with one agent that kept "talking like a pirate" in Discord while behaving correctly on Twitter.
- **Docker Configuration**: Users discussed challenges with Docker configuration for properly loading custom characters in Web UI.
- **Twitter Integration**: Questions about running Twitter agents on 1.0.0 beta and automating Twitter posts across multiple accounts with the same ElizaOS files.

### Creative Applications
- **AI Content Generation**: Jin demonstrated using Claude to analyze DAO weekly data and create Star Wars Jedi Council-inspired scripts based on character descriptions, noting the process took approximately 15 minutes.
- **Automated Social Media**: Hubert is developing an automated system for AI16Z news that pulls updates from a JSON file (updated daily at midnight UTC) and creates formatted posts on X using Puppeteer for screenshots, cron jobs for automation, and the X API for posting.

### Community & Business
- **Tokenomics Discussion**: Partners debated whether they should receive free token allocations or be required to invest, with suggestions for a stake-weight-reputation system where partners could stake tokens to vouch for projects.
- **Scam Alert**: A fraudulent ElizaOS token on BSC was identified and reported to the community.

## Key Questions & Answers

### Technical Questions
- **Q**: Is it possible to stream agent responses like how you see Claude or ChatGPT "typing" a response? (asked by scooper)  
  **A**: Unanswered, but identified as a desired feature

- **Q**: How can I handle user inputs in prompts if I want to swap ETH but didn't enter the recipient address? (asked by Waqas Wahid)  
  **A**: The AI can't pull from memory or past chats, so each transaction needs the address upfront. (answered by Willie)

- **Q**: How to define the saveProgress() function? (asked by Kimani)  
  **A**: Make it take parameters like currentStep and totalSteps, then save to a file or DB. (answered by Willie)

- **Q**: Is the JSON updated daily at a specific time? (asked by hubert)  
  **A**: Updates daily at midnight UTC (answered by jin)

### ElizaOS Questions
- **Q**: I wonder how smarter Spartan in V2 than a V1 (asked by Void)  
  **A**: About the same right now, the LLM model isn't changing, just how it experiences the world (answered by Odilitime)

- **Q**: Are packages from v1 coming to v2, including other LLM providers? (asked by bloxandewoks)  
  **A**: User was directed to ask in another channel after getting dev role (answered by Patt)

- **Q**: How long did it take to set up and get results from Claude for the Star Wars Jedi Council scene ideas? (asked by jin)  
  **A**: About 15 minutes, mostly spent gathering and creating character descriptions (answered by jin)

## Community Help & Collaboration

### Installation & Setup Help
- **Helper**: lefrog | **Helpee**: Pedro  
  **Context**: Error when starting V2 CLI  
  **Resolution**: Suggested adding @beta after the cli word in the command

- **Helper**: Pedro | **Helpee**: lefrog  
  **Context**: Installation errors with V2 CLI  
  **Resolution**: Shared solution to remove node_modules, package-lock.json and reinstall with bun

- **Helper**: Creed | **Helpee**: LuZ42  
  **Context**: Node version error  
  **Resolution**: Suggested using "nvm 23.3.0" command

### Agent Configuration Help
- **Helper**: jin | **Helpee**: bobo bixby  
  **Context**: Agent kept talking like a pirate despite configuration changes  
  **Resolution**: Suggested adding "no emojis" to character file and iteratively fixing character file

### Project Development Help
- **Helper**: jin | **Helpee**: hubert  
  **Context**: Automated social media posting system development  
  **Resolution**: Provided JSON update schedule and suggested adding a filter/editing pass using HackMD API

## Action Items

### Technical
- Fix plugin-sql installation errors in V2 CLI (Mentioned by jin)
- Fix connection to lifi in evm plugin for transfer/bridge/swap functionality (Mentioned by shadows.13)
- Resolve issue with character persistence in Discord vs Twitter (Mentioned by shadows.13)
- Fix Docker configuration to properly load custom characters in Web UI (Mentioned by shadows.13)
- Add extended thinking for Anthropic (Mentioned by shadows.13)
- Fix web interface settings access (Mentioned by shadows.13)
- Address the BSC scam token impersonating ElizaOS (Mentioned by Zolo)
- Prepare for v2 launch, launchpad, and tokenomics (Mentioned by DannyNOR NoFapArc)
- Implement daily X post system that pulls new updates from JSON (Mentioned by hubert)
- Add filter/editing pass before publishing social media updates (Mentioned by jin)
- Implement multi-agent communication functionality (Mentioned by Nice Dreams Maybe)
- Clarify LLM compatibility requirements with Eliza OS (Mentioned by winded4752)

### Documentation
- Create guide for setting up ElizaOS on Windows (Mentioned by Nice Dreams Maybe)
- Document how to run agents with multiple Twitter accounts (Mentioned by Creed)
- Create shorter tutorial videos for better user engagement (Mentioned by Nice Dreams Maybe)
- Create readable graphics for posts to avoid direct links (Mentioned by hubert)
- Add blog posts to eliza.how/blog (Mentioned by jin)

### Feature
- Add ability to stream agent responses like ChatGPT/Claude (Mentioned by scooper)
- Add support for Telegram message deletion (Mentioned by SergioAguilar)
- Implement stake-weight-reputation system for partners (Mentioned by yikesawjeez)
- Consider raffle mechanism for token allocation (Mentioned by Odilitime)
- Evaluate token distribution models for partners (Mentioned by DorianD)
- Expand package list in v2 to match v1 capabilities and add more LLM providers (Mentioned by bloxandewoks)
- Create canonical character profile descriptions file for future AI prompting (Mentioned by jin)
- Generate AI prompts for storyboard images (Mentioned by jin)
- Replace human assistant with AI solution (Mentioned by Patt)