Coverage
AI coverage
Public-source coverage focused on model, policy, and platform narratives; designed for signal-finding, not hype.
Source types (counts)
- RSS feeds: 41
- Reddit communities: 26
- Telegram channels: 9
- X accounts / queries: 55
- GitHub sources: 23
- Polymarket packs: 0
Counts reflect configured sources across enabled packs. We intentionally do not publish the full list.
Update cadence
- Ingest tick: about every 10m
- Base per-source minimum: 10m
- GitHub minimum: 60m
- X minimum: 12h (windows/day: 2)
- Reddit poll interval: 60m
- Polymarket poll: Pack-defined
Cadence is best-effort and batch-based, not real-time.
Coverage categories
- Model releases and benchmarks
- Policy and regulation
- Platforms and tooling
- Research and preprints
- Community amplification (limited)
Enabled pack groups: AI (5) · Mainstream (1)
Example publishers (not exhaustive)
Research
arXiv · bioRxiv · OpenAI
Platforms
Hugging Face · GitHub · PyPI
Policy
SEC · EU institutions · US agencies
What we intentionally do not include
- No internal company roadmaps or non-public papers.
- No private datasets or paid analyst terminals.
- No scraping behind logins.
Examples are illustrative only and not exhaustive. We intentionally do not publish the full source list.