文件历史

提交图

8 次代码提交

作者 SHA1 备注 提交日期
Alex Rudenko 482a288c7e chore: evaluate select_page scenario (#925)
Passes the test with gemini-2.5-flash with no errors.

Closes https://github.com/ChromeDevTools/chrome-devtools-mcp/issues/921
2026-02-10 10:11:23 +00:00
Alex Rudenko 4b8e9f2875 feat: allow skipping snapshot generation for input tools (#821)
Input tools have snapshots not re-generated by default now with an
option to opt-in to get a snapshot.

Refs: https://github.com/ChromeDevTools/chrome-devtools-mcp/issues/726
2026-01-23 13:43:00 +00:00
Alex Rudenko a8169676f9 feat: support device viewport and user agent emulation (#798)
This PR adds two emulation attributes `viewport` and `userAgent` that
allow emulating a mobile or a different desktop device.

Closes https://github.com/ChromeDevTools/chrome-devtools-mcp/pull/272
Closes https://github.com/ChromeDevTools/chrome-devtools-mcp/issues/619
Closes https://github.com/ChromeDevTools/chrome-devtools-mcp/issues/410
Closes https://github.com/ChromeDevTools/chrome-devtools-mcp/issues/280
(the recommended way is to emulate via the MCP server instead of a
parallel DevTools session)
2026-01-20 14:07:57 +00:00
Alex Rudenko aa9a176956 fix: improve performance tool description (#800)
Drive-by: refactor and improve eval scripts.
2026-01-20 11:38:51 +00:00
zyzyzyryxy f028f8c075 chore: Update eslint rule to require empty line after license notice. (#784)
We should make sure there is always an empty line after license notice,
otherwise it gets dropped during build if it happens to be before import
statement.

---------

Co-authored-by: Piotr Paulski <piotrpaulski@chromium.org>
2026-01-16 16:35:45 +00:00
Alex Rudenko 303a1383e1 chore: add more eval scenarios (#780) 2026-01-16 08:38:45 +00:00
Alex Rudenko 0284d18b7b chore: support an html route in evals (#770)
Allows using the server URL in the prompt.
2026-01-14 17:11:53 +00:00
Alex Rudenko 257b9943e9 chore: add basic eval (#766)
This PR adds a tool based on the node test runner that runs a loop to
see what tools a model chooses given a prompt. The expectations are
encoding for each prompt. Run `npm run eval` to get results. Currently,
only Gemini and needs and API key.
2026-01-14 14:15:21 +00:00