Buglyst Hiring · See what your reviewer receives
Review the work, then discuss it.
Refund issued twice after a webhook retry. This fictional standard assessment uses the same report component as the employer workspace.
Buglyst Hiring · Illustrative sample · Fictional work record, not a real candidate.
Sample Candidate (illustrative)
sample-candidate@example.invalid
Backend engineer · retry handling · Assessment v1
Some exercises have not yet been tested with enough candidate attempts. Use the results alongside an interview.
- Score
- 100%
- Required score
- 70% · reached
- Exercises passed
- 1/1
- Time spent
- 31m 7s
- Test runs
- 3
- Solutions submitted
- 1
Exercise results
Open an exercise to see the candidate's approach, changes, and test results.
Exercise 131m 7s · 3 test runs · 1 solution submittedPassed
Exercise version 1 · Not yet tested with enough candidate attempts
- Files explored
- Opened 4 files while investigating.
- What they thought was wrong
- Recorded 3 notes about the likely problem before submitting a solution.
- Testing
- Ran the tests 3 times and submitted 1 solution.
- Files changed
- Changed 1 file.
- Earlier attempts
- No submitted solution failed the tests.
- Final test results
- The final solution passed the additional tests that candidates cannot see.
Candidate steps and testing habits
- Opened README.md
- Ran tests · did not pass
- Opened src/webhooks/refund-handler.ts
- Opened src/refunds/service.ts
- Explanation: The queue can deliver the webhook more than once. I will check how the handler identifies a refund across deliveries.
- Explanation: The idempotency key includes delivery time, so a retry gets a different key. The refund ID stays the same across deliveries.
- Ran tests · passed
- Opened src/idempotency/keys.ts
- Ran tests · passed
- Explanation: I used the refund ID as the stable key, then reran the visible checks. I would also ask about concurrent deliveries before changing the idempotency store.
- Submitted a solution · passed
How to read these results
What to consider before deciding
These details help you judge the results. They do not prove misconduct or determine a hiring decision.
- Exercise access
- A private exercise, available to invited candidates
- How often this exercise has been used
- Infrequently
- Recorded approach
- No unusual pattern found
- Similarity to other solutions
- No similar solutions found
- Additional test results
- Every exercise has a recorded pass for the additional tests
- Some exercises have not yet been tested with enough candidate attempts. Use the results alongside an interview.
Report format: version 2. Assessment version 1.
AI use: what the candidate shared
This section separates what the candidate told us from what Buglyst recorded.
Used Buglyst assistant
- How it was used
- Asked about retry behavior and where to look for the idempotency key.
- How suggestions were checked
- Checked the handler and reran visible checks after changing the key.
- Suggestions they chose not to use
- Not recorded.
Buglyst recorded 2 requests to the built-in AI assistant and 3 test runs. Of the exercises that passed, 1 included a test run. This does not show whether AI caused a solution to pass.
This report shows the candidate's assessment results. Use it alongside interviews and other work samples. Your team makes the hiring decision.
Bring the evidence into your interview
These are human follow-up questions, not Buglyst’s recommendation about a person.
- Ask the candidate to explain why a delivery-time key changes across retries, using their recorded explanation.
- Discuss which visible checks failed and what the passing runs establish. Ask what concurrent-delivery behavior remains untested.
- Review the candidate’s AI disclosure separately from the recorded assistant-request count.
In a real completed candidate report, authorized reviewers can save a private debrief about evidence usefulness and interview follow-through. This sample does not save decisions or create attempts.
Available report detail follows the assessment’s disclosure settings. Missing records are labelled. Company-project reports do not currently capture file navigation, and company execution remains a controlled pilot pending production runner verification.
Choose a reviewed debugging assessment for your next engineering interview.
ILLUSTRATIVE SAMPLE · This is a demonstration of Buglyst’s report structure. It does not represent a real candidate. · buglyst.com/hiring/sample-report