Research the following and write an incident report using the template in ./.claude/skills/research-item/template.md.

Steps

1 — Fetch input URLs

Use this fallback chain — stop at the first that returns usable content:

WebFetch <url> — preferred; returns clean text with no local cost
./bin/fetch-url --extract <url> — local fallback using trafilatura; use when WebFetch fails or is blocked
./bin/fetch-url --og <url> — last resort; returns only title/description/url (~50 tokens); enough for triage if full content is unavailable
./bin/fetch-url <url> (raw HTML) — Threads URLs only; extract post text with:
```
grep -o '"text":"[^"]*"' <saved-file> | head -20
```

Exit code 2 from --extract means trafilatura got no content (JS-rendered or paywalled) — fall back to --og.

Research the following and write an incident report using the template in ./.claude/skills/research-item/template.md.

1 — Fetch input URLs

Use this fallback chain — stop at the first that returns usable content:

WebFetch <url> — preferred; returns clean text with no local cost
./bin/fetch-url --extract <url> — local fallback using trafilatura; use when WebFetch fails or is blocked
./bin/fetch-url --og <url> — last resort; returns only title/description/url (~50 tokens); enough for triage if full content is unavailable
./bin/fetch-url <url> (raw HTML) — Threads URLs only; extract post text with:
```
grep -o '"text":"[^"]*"' <saved-file> | head -20
```

Exit code 2 from --extract means trafilatura got no content (JS-rendered or paywalled) — fall back to --og.

Input type	Action
Single incident	Research → create/update file
List/roundup	Split into individual incidents, research each
Statistical claim	Verify against history.md
Pre-2026-04-07 with no direct link	Skip
No news coverage found	INDEX entry only, no file

Research Item