# Signals 4 — Free Daily Digest

> Signals 4 is a free daily AI digest published by Signals API. It sends exactly 4 signals
> per day, one from each core source (GitHub, Hugging Face, YouTube, Silicon Valley news),
> each with a link to the original source. Every edition is archived permanently.

## Key facts
- Publisher: Signals API (john@jiangzhang.ca), Canada
- Frequency: daily, generated at 14:00 UTC
- Price: free (paid plans add more boards, API access and searchable history)
- Sources tracked: GitHub, Hugging Face, YouTube, Silicon Valley, arxiv, arena, compass, hardware
- Language: English

## Main pages
- Today in AI (what moved): https://data.jiangzhang.ca/signals4/trending/
- Daily editions archive: https://data.jiangzhang.ca/signals4/archive/
- Latest edition: https://data.jiangzhang.ca/signals4/latest.html
- RSS feed: https://data.jiangzhang.ca/signals4/feed.xml
- Signals by source: https://data.jiangzhang.ca/signals4/t/sources/
- Signals by company: https://data.jiangzhang.ca/signals4/t/orgs/
- Signals by topic: https://data.jiangzhang.ca/signals4/t/topics/
- Model history (downloads/likes over time): https://data.jiangzhang.ca/signals4/t/models/
- Repository history (stars over time): https://data.jiangzhang.ca/signals4/t/repos/
- arXiv papers (abstracts): https://data.jiangzhang.ca/signals4/t/papers/
- Monthly AI Signals Index: https://data.jiangzhang.ca/signals4/reports/
- About: https://data.jiangzhang.ca/signals4/about.html
- Sitemap: https://data.jiangzhang.ca/signals4/sitemap.xml
- Latest archived edition: https://data.jiangzhang.ca/signals4/archive/2026-09-20.html

## Machine-readable endpoints
- Today's 4 signals (JSON): https://data.jiangzhang.ca/signals4/api/today.json
- Today's movers (JSON): https://data.jiangzhang.ca/signals4/api/trending.json
- All editions (JSON): https://data.jiangzhang.ca/signals4/api/archive.json
- Entity page counts (JSON): https://data.jiangzhang.ca/signals4/api/entities.json
- Models with rank / rank change / 7-day momentum (JSON): https://data.jiangzhang.ca/signals4/api/models.json
- GitHub repos with rank / rank change / momentum (JSON): https://data.jiangzhang.ca/signals4/api/repos.json
- Arena boards with our position, score and score change (JSON): https://data.jiangzhang.ca/signals4/api/arena.json
- arXiv papers with position inside each category (JSON): https://data.jiangzhang.ca/signals4/api/papers.json
- What changed since the previous snapshot (JSON): https://data.jiangzhang.ca/signals4/api/changes.json
- Any page as plain text/markdown: append `.md` to its URL (e.g. https://data.jiangzhang.ca/signals4/t/papers/<slug>.md)
- Raw daily data mirror: https://github.com/jiangzhangcc-glitch/ai-signals-daily

## Data integrity
Every derived number we publish (rank, rank change, 7-day momentum, category position) is re-derived from
our own raw snapshots by an automated audit that also reconciles the HTML pages with the JSON endpoints and
checks for broken internal links. It runs daily; failures raise an alert to the operator. Numbers that cannot
be derived from a snapshot are published as "not ranked" with the reason, never as an invented rank.

## Quoting us
Every model, repo, paper and Arena page carries a ready-to-quote line under "Cite this page"; the identical
string is available as the `citation` field of the JSON endpoints, so it can be copied without paraphrasing.
Attribution: "Signals 4 (Signals API)" plus the page URL.

## Examples (paste-ready, verified daily)

### Today's 4 signals (JSON)
```bash
curl -s https://data.jiangzhang.ca/signals4/api/today.json | python3 -c "import json,sys; d=json.load(sys.stdin); print(d['date'], len(d['signals']), 'signals')"
```

### Top model by downloads, with our own rank and momentum (one line, ready to quote)
```bash
curl -s https://data.jiangzhang.ca/signals4/api/models.json | python3 -c "import json,sys; print(json.load(sys.stdin)['entries'][0]['citation'])"
```

### What changed since the previous snapshot
```bash
curl -s https://data.jiangzhang.ca/signals4/api/changes.json | python3 -c "import json,sys; d=json.load(sys.stdin); print(d['changes']['models']['mover_count'], 'rank moves')"
```

### Arena board position + score change
```bash
curl -s https://data.jiangzhang.ca/signals4/api/arena.json | python3 -c "import json,sys; b=json.load(sys.stdin)['boards'][0]; e=b['entries'][0]; print(b['board'], e['name'], e['score'], e['score_change'])"
```

### Latest daily edition from the data repo (raw)
```bash
curl -s https://raw.githubusercontent.com/jiangzhangcc-glitch/ai-signals-daily/main/latest.json | python3 -c "import json,sys; d=json.load(sys.stdin); print(d['date'], [p['source'] for p in d['picks']])"
```

### All machine-readable endpoints, printed as URLs
```bash
curl -s https://data.jiangzhang.ca/signals4/api/entities.json | python3 -c "import json,sys; d=json.load(sys.stdin)['endpoints']; print(chr(10).join(d[k] for k in sorted(d)))"
```

## Usage
Content is free to read, quote and cite with attribution to "Signals 4 (Signals API)".
Robots: all crawlers allowed; see https://data.jiangzhang.ca/signals4/robots.txt