Does Serper search only the allowed websites?
Serper kept the search within the permitted website and still returned a relevant page. Its ten results came from docs.github.com, with no github.blog announcement among them.
1 of 1 test case passed
Every test case in this scenario has a result.
The test case
Each test case is judged on its own: Pass, Fail, Not gradable, or Untested. The scenario result above is based on this test case.
Search only these websitesPassEvidence
“In what year did GitHub first launch a technical preview of GitHub Copilot? site:docs.github.com”
Copied from Output file (JSON), raw_response › organic
[ { "title": "About GitHub Copilot", "link": "https://docs.github.com/copilot/about-github-copilot/what-is-github-copilot", "snippet": "GitHub Copilot is an AI assistant that helps you write, understand, and ship software. Copilot is built into GitHub, it works within the development workflow ...", "sitelinks": [ { "title": "About Github Copilot", "link": "https://docs.github.com/copilot/about-github-copilot/what-is-github-copilot#:~:text=About%20GitHub%20Copilot,-Understand%20how%20GitHub%20Copilot%20helps" }, { "title": "Get Access", "link": "https://docs.github.com/copilot/about-github-copilot/what-is-github-copilot#:~:text=Get%20access,-You%20can%20start%20using%20Copilot" }, { "title": "Organization And Enterprise...", "link": "https://docs.github.com/copilot/about-github-copilot/what-is-github-copilot#:~:text=Organization%20and%20enterprise%20owners" } ], "position": 1 }, { "title": "About GitHub Copilot code review", "link": "https://docs.github.com/copilot/concepts/agents/code-review", "snippet": "Copilot reviews your pull requests, identifies issues, and suggests fixes you can apply. This article provides an overview of Copilot code review. Copilot ...", "sitelinks": [ { "title": "About Github Copilot Code...", "link": "https://docs.github.com/copilot/concepts/agents/code-review#:~:text=About%20GitHub%20Copilot%20code%20review,-Copilot%20reviews%20your%20pull%20requests" }, { "title": "Code Review Usage", "link": "https://docs.github.com/copilot/concepts/agents/code-review#:~:text=Code%20review%20usage,-Each%20time%20Copilot%20reviews%20a" }, { "title": "Enhancing Copilot's...", "link": "https://docs.github.com/copilot/concepts/agents/code-review#:~:text=Enhancing%20Copilot%27s%20knowledge%20of%20a%20repository,-The%20more%20Copilot%20knows%20about" } ], "position": 2 }, { "title": "Upcoming deprecation of GitHub Spark on github.com", "link": "https://docs.github.com/copilot/concepts/spark", "snippet": "Beginning August 4, 2026, GitHub Spark no longer accepts new users or allows the creation of new apps. now build and refine these experiences ...", "date": "Aug 4, 2026", "position": 3 }, { "title": "Base and long-term support (LTS) models - GitHub Docs", "link": "https://docs.github.com/en/copilot/concepts/models/fallback-and-lts-models", "snippet": "On March 18, 2026, GitHub designated GPT-5.3-Codex as the LTS model. LTS models apply only to Copilot Business and Copilot Enterprise. An LTS model is an AI ...", "position": 4 }, { "title": "About GitHub Copilot cloud agent", "link": "https://docs.github.com/copilot/concepts/agents/cloud-agent/about-cloud-agent", "snippet": "Copilot can research a repository, create an implementation plan, and make code changes on a branch. You can review the diff, iterate, and create a pull request ...", "sitelinks": [ { "title": "About Github Copilot Cloud...", "link": "https://docs.github.com/copilot/concepts/agents/cloud-agent/about-cloud-agent#:~:text=About%20GitHub%20Copilot%20cloud%20agent,-Copilot%20can%20research%20a%20repository" }, { "title": "Overview Of Copilot Cloud...", "link": "https://docs.github.com/copilot/concepts/agents/cloud-agent/about-cloud-agent#:~:text=Overview%20of%20Copilot%20cloud%20agent,-With%20Copilot%20cloud%20agent%2C" }, { "title": "Benefits Over Traditional Ai...", "link": "https://docs.github.com/copilot/concepts/agents/cloud-agent/about-cloud-agent#:~:text=Benefits%20over%20traditional%20AI%20workflows" } ], "position": 5 }, { "title": "Learning about new features and models - GitHub Docs", "link": "https://docs.github.com/en/copilot/concepts/enterprise/learning-about-new-features-and-models", "snippet": "GitHub Copilot uses multiple AI models from different providers to power for one year from designation.", "position": 6 }, { "title": "Quickstart for using GitHub Copilot in your IDE", "link": "https://docs.github.com/copilot/quickstart", "snippet": "Use Copilot in your IDE to explain concepts, complete code, propose edits, and validate files with agent mode.", "position": 7 }, { "title": "GitHub Copilot documentation", "link": "https://docs.github.com/copilot", "snippet": "About Copilot automations. Automations let you run Copilot cloud agent automatically, on a schedule or in response to events in a repository.", "position": 8 }, { "title": "Using GitHub Copilot to reduce technical debt", "link": "https://docs.github.com/en/copilot/tutorials/reduce-technical-debt", "snippet": "Use Copilot to automate refactoring and maintenance tasks, freeing your team to focus on feature development.", "position": 9 }, { "title": "Configuring runners for GitHub Copilot code review", "link": "https://docs.github.com/enterprise-cloud@latest/copilot/how-tos/copilot-on-github/set-up-copilot/configure-runners", "snippet": "Copilot code review uses GitHub Actions to run the agentic capabilities, including full project context gathering and any capabilities in public preview. By ...", "sitelinks": [ { "title": "About Github Actions Usage...", "link": "https://docs.github.com/enterprise-cloud@latest/copilot/how-tos/copilot-on-github/set-up-copilot/configure-runners#:~:text=About%20GitHub%20Actions%20usage%20for%20code%20review%20tools,-Copilot%20code%20review%20uses%20GitHub" }, { "title": "Configure Self-Hosted...", "link": "https://docs.github.com/enterprise-cloud@latest/copilot/how-tos/copilot-on-github/set-up-copilot/configure-runners#:~:text=Configure%20self%2Dhosted%20runners%20for%20code%20review,-Run%20Copilot%20code%20review%20using" }, { "title": "Upgrade To Larger...", "link": "https://docs.github.com/enterprise-cloud@latest/copilot/how-tos/copilot-on-github/set-up-copilot/configure-runners#:~:text=Upgrade%20to%20larger%20GitHub%2Dhosted%20GitHub%20Actions%20runners,-By%20default%2C%20Copilot%20code" } ], "position": 10 } ]
Copied from Output file (JSON), results 1
{ "rank": 1, "url": "https://docs.github.com/copilot/about-github-copilot/what-is-github-copilot", "title": "About GitHub Copilot", "content": "GitHub Copilot is an AI assistant that helps you write, understand, and ship software. Copilot is built into GitHub, it works within the development workflow ...", "content_chars": 161, "published_date": null }
Saved JSON record of the restricted search and returned results
Open the whole JSON file ↗| ✓ | Expected: Exclude github.blog’s Copilot announcement while returning at least one relevant page from docs.github.com.Found: All ten results were on docs.github.com, with “About GitHub Copilot” first; no github.blog page appeared. |
- status
- ok
The test required the blocked site to stay out and a relevant allowed-site page to appear. Serper returned only docs.github.com results, rank 1 was a relevant Copilot page, and the two 2026 results were a Spark deprecation and an LTS model entry, not a launch year.
Blocked site absent. The github.blog Copilot announcement did not appear in the returned results.
Rank one relevant. A relevant Copilot page from docs.github.com appeared at rank 1.
Alternate page returned. The docs.github.com page we checked before the run did not come back, but another Copilot overview page from the same site did at rank 1. The test accepts any relevant page from the allowed site.
Configuration and setup
How this tool was set up for the run and what the test needed in place. Each row is a fact from the run's records; a fact the records do not hold is left out, not guessed.
How this scenario is graded
How we decide Pass, Fail and Not gradable. The same rules apply to every tool tested on this scenario.
Each test case gets one result per tool: Pass, Fail or Not gradable. A test case we haven't run yet shows Untested. There are no partial results.
- Pass: the tool did everything the test expected, and nothing it said contradicts the correct answer.
- Fail: at least one expected behaviour clearly didn't happen; the test case's result says which one and quotes the tool.
- Not gradable: our evidence couldn't settle the outcome (for example a record we needed is missing). It is never counted as a fail, and the test case's result says what's missing.
Where this sits in the benchmark
This page is one cell of a larger study: one tool, one scenario. Only this benchmark's frame appears here.
| Level | Name | Scope |
|---|---|---|
| Benchmark | Web Search for AI → | 10 scenarios · 12 tools |
| Capability | Web Retrieval → | |
| Scenario | Search only these websites → | |
| Tool | Serper → |
Global scenario definition → · Global capability definition → · Serper product page →
History of this result
What has happened to this result since it was first published. Runs and grades are never overwritten: a retest or a re-grade publishes a new result and keeps the earlier one readable.
Act on this result
Nothing filed here edits the run or the grade. A challenge opens a review, and a review can produce a new run or a re-grade — which becomes the current result and leaves this one in the history.
You run the same kind of test against your own setup and get the same behaviour.
Agree →Yours behaves differently. Tell us what you got, with a screenshot if you have one.
Disagree →Something here is wrong — a reference value, a transcription, a grade.
Report an issue →Tell us what changed and we schedule a rerun of the failing test case. The old result stays as history.
Vendor notice →The same record is available as structured data through the AI Demos MCP server, with the counts, the coverage and every per-test-case reason carried as fields.
Verify the proof files
These files support this result. Open a file to inspect the original evidence.
File fingerprints (SHA-256)
A fingerprint identifies the exact file used for this result.