name: digest
description: >
Generate a multi-platform activity digest for a topic. Fetches and ranks
items from HN, GitHub, Reddit, YouTube, ethresear.ch, Snapshot, Polymarket,
package registries, CoinGecko, Blockscout, and Shodan. TRIGGER when: user
invokes "/digest" or asks for a "digest", "what's happening with X",
"activity summary for X", "news about X".
DO NOT TRIGGER when: user asks about digest agent code/implementation.
metadata:
author: DROOdotFOO
version: "1.0.0"
tags: digest, research, monitoring, news
argument-hint: '"" [--days N] [--platforms p1,p2]'
Digest
Generate a synthesized activity digest for any topic across multiple platforms.
What You Get
- Ranked list of items across platforms, weighted by engagement and recency
- Claude-synthesized narrative with citations (or raw items with
--no-synthesis)
- Markdown file or terminal output
Usage
digest generate "rust async runtime"
digest generate "zero knowledge proofs" --days 7 --platforms hn,github,reddit
digest generate "noir language" --output digest.md
digest generate "elixir otp" --no-synthesis
digest generate "noir" --no-expansion
digest list-platforms
Available Platforms
| Platform |
Key |
Signal source |
| Hacker News |
hn |
Algolia search (points, comments) |
| GitHub |
github |
gh CLI search (stars, forks) |
| Reddit |
reddit |
Search JSON API (upvotes, comments) |
| YouTube |
youtube |
yt-dlp flat-playlist (views, likes) |
| ethresear.ch |
ethresearch |
Discourse search (views, likes, posts) |
| Snapshot |
snapshot |
GraphQL governance (votes, scores) |
| Polymarket |
polymarket |
Gamma API (volume traded) |
| Packages |
packages |
hex.pm + crates.io + npm (downloads) |
| CoinGecko |
coingecko |
Trending, gainers/losers, new listings |
| Blockscout |
blockscout |
On-chain transfers, address activity |
| Shodan |
shodan |
Hosts, services, CVEs (SHODAN_API_KEY) |
Flags
| Flag |
Default |
Description |
--days / -d |
30 |
Lookback window in days |
--platforms / -p |
hn,github |
Comma-separated platform list |
--max-items / -n |
50 |
Max items per platform |
--output / -o |
(none) |
Write markdown to file |
--no-synthesis |
false |
Skip Claude narrative |
--no-expansion |
false |
Skip query expansion |
Pipeline
- Query expansion (static rules, LLM fallback planned)
- Parallel fetch across selected adapters
- Dedupe by URL + title similarity
- Rank by log-weighted engagement + recency decay
- Synthesize narrative via Claude (unless
--no-synthesis)
- Output to terminal (rich) or markdown file
Reference
By