Add serpdive extension

[edendalexis]

• I've read CONTRIBUTING.md and followed the relevant guidance for adding or updating my extension.

Adds SERPdive as a context server extension: web search for AI agents that returns the extracted, answer-ready content of each source page instead of a list of links, cleaned and sized for a context window. It wraps the serpdive-mcp npm package over stdio, in the same shape as the existing mcp-server-tavily extension.

Measured, not asserted: on a public, replayable 1,000-question benchmark judged blind by an independent model, SERPdive runs at the same speed as Tavily, feeds the LLM 20.2% fewer tokens, and wins 60.7% of decided quality duels against Tavily's default search — methodology and per-question results at https://github.com/edendalexis/serpdive-benchmark.

Extension repository: https://github.com/serpdive/zed-serpdive-mcp (MIT). One tool, serpdive_search, with two retrieval depths (fast key sentences, or full page content) and an optional written answer built from the sources. Free tier: 1,000 credits/month, no card.

Tested locally as a dev extension, per CONTRIBUTING:

• Builds clean for wasm32-wasip2 (the target this repo's CI uses) with Rust 1.97.
• Installed into Zed 1.11.3 as a dev extension: the extension host loaded it, ran context_server_command, installed serpdive-mcp@0.1.5 into the extension work directory, and started the server — node .../extensions/work/serpdive/node_modules/serpdive-mcp/bin/serpdive-mcp.js was running with the key from settings.
• Drove the underlying MCP server through the same invocation the extension uses: initialize returns the server info, tools/list returns serpdive_search.
• Missing-key path returns the intended error rather than panicking.

In this repo: pnpm sort-extensions (both extensions.toml and .gitmodules are sorted and the sorter is a no-op on re-run), pnpm build, and pnpm test (111 tests) all pass locally.

A key is created at https://serpdive.com/dashboard/keys and the free tier needs no card.

[cla-bot]

We require contributors to sign our Contributor License Agreement, and we don't have @edendalexis on file. You can sign our CLA at https://zed.dev/cla. Once you've signed, post a comment here that says '@cla-bot check'.

[cla-bot]

We require contributors to sign our Contributor License Agreement, and we don't have @edendalexis on file. You can sign our CLA at https://zed.dev/cla. Once you've signed, post a comment here that says '@cla-bot check'.

[edendalexis]

@cla-bot check

[cla-bot]

The cla-bot has been summoned, and re-checked this pull request!

[edendalexis]

Update while this is open: SERPdive now has a free tier, which changes the calculus for this integration.

model: "krill" is free and unlimited under fair use — no card, no credits to spend, no trial clock. It returns about 700 tokens per search (roughly half what the usual alternatives send), one request at a time, at low priority. The paid models are unchanged.

Why it matters here: the usual objection to adding a search provider is that reviewers and users need a paid key just to try it. That objection is gone — anyone can run this integration end to end for free, including whoever reviews this PR.

Nothing in this PR needs to change for that: the default stays mako, and krill is opt-in via the existing model parameter. Happy to add a line to the docs here if you'd like the free tier called out explicitly.