#15530: docs(web_fetch): document markdown-first Accept header and cf-markdown extraction

Docs follow-up for recent web_fetch markdown handling changes. This updates docs to clarify: - web_fetch sends markdown-first Accept header: `text/markdown, text/html;q=0.9, */*;q=0.1` - text/markdown responses use the `cf-markdown` extraction path Updated files: - docs/tools/web.md - docs/tools/index.md No code changes; documentation-only PR. <!-- greptile_comment --> <h2>Greptile Overview</h2> <h3>Greptile Summary</h3> This PR is a documentation-only follow-up clarifying `web_fetch`’s markdown-first `Accept` header (`text/markdown, text/html;q=0.9, */*;q=0.1`) and that `text/markdown` responses use the `cf-markdown` extraction path. The docs match the implementation in `src/agents/tools/web-fetch.ts` for the `Accept` header and `text/markdown` handling. One note in `docs/tools/web.md` overstates when `web_fetch` errors: for other content types the tool can return the raw body (`extractor: "raw"`) rather than failing extraction. <h3>Confidence Score: 4/5</h3> - This PR is safe to merge after a small docs precision fix. - Changes are documentation-only and align with the current `web_fetch` implementation for the markdown-first Accept header and `cf-markdown` extractor. The only issue found is a slightly misleading error-condition sentence in the web_fetch notes section. - docs/tools/web.md _{Last reviewed commit: 7339d66} <!-- greptile_other_comments_section --> _{(2/5) Greptile learns from your feedback when you react with thumbs up/down!} <!-- /greptile_comment -->