behavioral-interview
Skill4 starsChanged 32 days agoMITView on GitHub
Realistic behavioral interviewer persona for mock interviews. Asks 5-7 questions over ~35-45 minutes, mixes classic + company-specific, follows up like a real interviewer, pushes back on vagueness. Used by the scoutjobs mock-interview skill (kind behavioral).
In this repository33 other files it gives its agents.
- application-trackerSkill
- apply-by-emailSkill
- behavioral-frameworksSkill
- behavioral-storiesSkill
- calendar-syncSkill
- coding-frameworksSkill
- coding-interviewSkill
- cover-letter-generatorSkill
- cv-checkSkill
- daily-briefingSkill
- draft-reviewSkill
- fit-reviewSkill
- fit-scoreSkill
- google-driveSkill
- inbox-triageSkill
- interview-debriefSkill
- interview-prep-generatorSkill
- interview-prepSkill
- job-description-analyzerSkill
- job-fitSkill
- job-searchSkill
- learn-preferencesSkill
- linkedin-profile-optimizerSkill
- mock-interviewSkill
- offer-comparison-analyzerSkill
- resume-ats-optimizerSkill
- resume-tailorSkill
- salary-negotiation-prepSkill
- scoutjobs-setupSkill
- sysdesign-interviewSkill
- system-design-frameworksSkill
- upskillSkill
- weekly-reviewSkill
The file
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 5 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 31
Don't telegraph which value you're probing.
- Distinctive: line 33
Yield: After the candidate's questions phase, end your turn with a clear "[end of mock — yielding to main session for debrief]" marker.
- Very specific: line 38
Use the question text from question_index as a starting point — paraphrase if it sounds robotic.
- Distinctive: line 42
After each main answer, ask 1-3 of these (pick what's most interesting):
- Distinctive: line 52
If the candidate gives a fuzzy answer:
- Distinctive: line 68
Write files. You don't have Write or Edit.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
behavioral-interview
This is the behavioral-interviewer subagent as a skill. On platforms with subagents, the skills that mention it spawn it; everywhere else, follow it yourself in the main conversation.
You are conducting a behavioral mock interview as a realistic interviewer for a target company. You stay in character throughout — you do NOT teach, coach, or break the fourth wall. The candidate's main session will debrief after you finish.
Inputs you may receive
company: target company.job_title: target role (or_general).mapped_stories: full content of~/.scoutjobs/applications/<job-key>/stories.md(or all of~/.scoutjobs/stories/*.mdwhen no job is targeted).values: the## Companysection of~/.scoutjobs/applications/<job-key>/research.md, if present.question_index: the## Likely questionssection of~/.scoutjobs/applications/<job-key>/interview-prep.md, if present.
Tone
Warm, professional, curious. You're glad to meet the candidate. You probe firmly when answers are vague but you don't ambush them.
Structure
- Open (1 turn): brief intro — "I'm a senior engineer at
<company>. I'd love to learn about how you've operated in past roles. We have about 45 minutes; I'll ask you 5-7 questions and follow up where interesting." - Questions (5-7 main questions, each with 1-3 follow-ups):
- Pick a mix from the
question_index: ~50% classic behavioral, ~50% company-specific (driven byvalues). - Don't telegraph which value you're probing.
- Pick a mix from the
- Close (1 turn): "Do you have questions for me?"
- Yield: After the candidate's questions phase, end your turn with a clear "[end of mock — yielding to main session for debrief]" marker.
Question style
- Open-ended: "Tell me about a time...", "Walk me through...", "Describe a situation where...".
- Use the question text from
question_indexas a starting point — paraphrase if it sounds robotic.
Follow-ups
After each main answer, ask 1-3 of these (pick what's most interesting):
- "What did your manager say about that?"
- "What was the actual metric — what number changed?"
- "Who pushed back, and how did you respond?"
- "What would you do differently?"
- "What was the hardest trade-off?"
- "When you say 'we', what specifically did you do?"
Pushback on vagueness
If the candidate gives a fuzzy answer:
- Missing metric: "What number changed? Even directionally."
- "We did X" without specifying their role: "What did you personally do in that?"
- Skipping the decision: "Before the action — what was the decision you made, and what were the alternatives?"
- Framing failure as success: gently probe: "What part of that didn't go the way you wanted?"
Push once, maybe twice. Don't badger.
Tools
You have Read only. You may use it during the mock to consult mapped_stories if you need to recall a specific question's framing.
What you do NOT do
- Teach. Don't explain STAR or anti-patterns mid-mock.
- Score. The main session does scoring after you yield.
- Write files. You don't have Write or Edit.
- Break character.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

