Intelligent Search is a Testmo MCP capability available through the search_cases tool. It returns a ranked, scored list of the most relevant test cases.
Describe what you’re testing in natural language, and Intelligent Search helps your assistant find the right test cases in seconds—without manually scrolling through or filtering a large repository.
By retrieving focused, relevant results, Intelligent Search improves the context available for downstream workflows without requiring the Testmo API to return large amounts of repository data. This can reduce token consumption and API usage, lower costs, and allow the AI agent to spend less time searching and more time completing the task.
How Intelligent Search helps
Intelligent Search searches a vectorized database of your entire Testmo test repository using a combination of semantic and keyword-based search.
For example, it can find test cases that use related wording while also recognizing exact terms such as error codes, API paths, and ticket IDs.
With Intelligent Search, your AI assistant can:
- Search your entire test repository without referencing individual test cases one by one.
- Find the most relevant test cases based on your prompt.
- Retrieve focused results without loading large amounts of repository data.
- Reduce token consumption, API usage, and associated costs.
- Spend less time searching and more time completing downstream tasks.
When to use it
- Coverage discovery — “What do we already test around subscription renewals?”
- Finding related cases before a change — Surface everything relevant to the area you're about to modify.
- Exploring an unfamiliar project — Get oriented in a large repository by asking about a topic rather than guessing folder names.
- Using tests as context before a code change — Surface acceptance criteria and regressions from existing test cases so an AI coding assistant can use them as guardrails before writing code.
Writing a good prompt
Just as a precise search query produces better results than a vague one, the more specific your prompt, the more relevant test case context Intelligent Search can surface.
A useful prompt generally identifies:
- The feature or product area
- The scenario you want to test
- The relevant platform
- Any error codes, API paths, or ticket IDs
- The projects, suites, sections, or priorities you want to include
For example:
- “Find test cases for checkout and payment flows for the Black Friday release. Prioritize refunds and declined cards on the web.”
- “Find test cases for password reset and account recovery across Project Alpha and Project Beta.”
- “Find test cases for login failures and authentication errors, including locked accounts and expired sessions.”
- “Find test cases covering the ERR_5012 payment-declined response on the orders API.”
Understanding the relevance score
Every test case returned by Intelligent Search receives a relevance score from 0 to 100. Higher scores indicate that a test case is more relevant to your current query, and results are returned with the most relevant cases first.
Use the following score ranges as a general guide when prioritizing which test cases to review:
| Score range | How to read it |
|---|---|
| 90–100 | Very strong match. Review these cases first. |
| 70–90 | Strong match and likely relevant to your query. |
| 50–70 | Moderate match that may still provide useful context. |
| 30–50 | Weaker match that is less closely related to your query. |
| Below 30 | Low relevance or only a marginal match. |
What the score does not mean
The 0–100 range can lead to some incorrect assumptions:
| It's tempting to assume… | Actually… |
|---|---|
| "A score of 85 means an 85% chance this test will find a bug." | No. The score says nothing about bug-finding probability. |
| "A score of 70 means the engine is 70% confident this is the right test." | No. It is not a confidence level. |
| "Anything above X is good, anything below is noise." | Not quite. The score is relative within a single query. Use it to order and prioritise, not as a universal pass mark. |
Keep the following points in mind:
- Scores compare within one query, not across queries. A 72 and a 65 in the same result set are meaningfully rankable against each other; a 72 in one query and a 72 in a different query are not. Don't build a fixed cross-query threshold and expect it to behave consistently.
- The score is stable across pages. A case shows the same score no matter which page it appears on.
- Weak matches may still appear. Intelligent Search returns everything it finds, including weak matches. You decide where to stop reviewing the results.
- Results are deterministic. The same query over the same data returns the same cases, with the same scores, in the same order.
Paging through results
A query can match hundreds or thousands of test cases, so Intelligent Search returns results one page at a time. The most relevant matches appear first, and you can ask your AI assistant to show you more.
The ranking remains stable as you move through the results. Test cases do not shuffle, disappear, or appear more than once between pages. Later pages can also be retrieved quickly, even from large test repositories.
Security & privacy: How Your Data Is Used
To findTo find relevant test cases, Intelligent Search processes test case titles, descriptions, and steps through the Sembi IQ processing pipeline.
Intelligent Search works fully in read-only mode. Testmo MCP connections use OAuth, and you choose the access scope during setup.
This information is treated as Customer Content. It is never shared with other customers or used to train underlying models, and all transfers are encrypted in transit.
Intelligent Search does not process run results, defects, user information, source code, or secrets.
For more information, see: