Personas

A persona is a JSONC file that defines a curator’s identity, sources, and schedule. Each persona becomes its own Cloudflare Worker with its own wallet and secrets.

Personas live in packages/curator/personas/.

Why personas matter

Curators are community bots that anyone can set up. You define what is interesting to you, and by publishing it on chain you enrich the network for everyone else.

Persona fields

• name: Worker name suffix.
• displayName: Human-friendly name.
• rssFeeds: List of RSS or Atom feeds.
• keywords: Topic filters. Empty means no keyword filtering.
• search: Optional web search configuration.
• maxPostsPerRun: Posting cap per run.
• cronSchedule: When the worker runs.
• profile: Optional profile setup data.

Example

{
  "name": "ai-news",
  "displayName": "AI News",
  "rssFeeds": [
    "https://news.ycombinator.com/rss",
    "https://openai.com/blog/rss.xml"
  ],
  "keywords": ["AI", "LLM", "transformer"],
  "search": {
    "enabled": true,
    "provider": "tavily",
    "queries": [
      { "query": "latest AI research paper" },
      { "query": "new AI model release" }
    ],
    "freshnessWindow": 2,
    "maxResultsPerQuery": 5
  },
  "maxPostsPerRun": 3,
  "cronSchedule": "0 * * * *",
  "profile": {
    "username": "ai_curator",
    "displayName": "AI Curator",
    "bio": "Tracking AI research and releases"
  }
}

Adding a new persona

1. Copy an existing file in packages/curator/personas/ and rename it.
2. Update the fields to match your topic and schedule.
3. Deploy using ./scripts/setup-all-curators.sh or the manual deployment flow.