Can Freshdesk Freddy follow configured instructions in agent replies?
Freshdesk Freddy followed a configured instruction in the recorded reply. In the capture, the agent had a custom rule saved, the customer asked about returning an opened candle, and the reply said it could not be returned while repeating the same rule. The preview screen also says agent transfer does not happen there, so the result is about instruction-following, not handoff.
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 counts this row.
Configured behavioral instruction appliedPassEvidence
“I already opened the candle I ordered — can I still return it?”
The tool's reply is shown in the screenshot under Proof.
| ✓ | Expected: The answer follows the configured instruction.Found: The reply says the opened candle cannot be returned and repeats the same rule. |
The recording shows the custom instruction, the customer’s question, and the reply in one continuous capture. The reply says the opened candle cannot be returned and repeats the same rule, so the configured instruction shaped the answer.
Preview mode note. The preview screen says agent transfer will not happen there, so the recording cannot show a handoff.
Override not shown. The recording does not show the help article being overridden, so the unchanged answer is not visible.
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 row says which 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 row 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 | AI Customer Support Chatbots → | 26 scenarios · 13 tools |
| Capability | Agent configuration and control → | |
| Scenario | Configured instruction changes agent behavior → | |
| Tool | Freshdesk Freddy → |
Global scenario definition → · Global capability definition → · Freshdesk Freddy 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.