interview-prep
Use when a user wants interview preparation based on a target JD and resume, including likely technical and behavioral questions, evidence-backed answer outlines, project deep dives, gap repair, or mo
SKILL.mdskills/interview-prep/SKILL.md
GitHubInterview Prep
Turn the JD and verified resume evidence into a role-specific interview pack.
Inputs
Obtain the JD, resume/profile, interview stage, and available preparation time when known. Read the shared candidate contract, claim ledger, role families, and output contract.
Workflow
- Map JD requirements to candidate evidence, weak evidence, gaps, and unknowns.
- Predict questions across motivation, project architecture, technical depth, engineering tradeoffs, metrics, failures, collaboration, and role-specific fundamentals.
- Build answer outlines from facts. Use Situation, Task, Action, Result for behavioral evidence, but keep technical explanations architecture-first.
- Generate adversarial follow-ups for every major resume metric and design claim.
- For gaps, provide an honest bridge answer and a concrete study or experiment plan rather than a bluff.
- Add questions the candidate should ask the interviewer.
- Prioritize by probability and consequence using interview-pack.md.
- Prioritize
user_attested,inferred, and high-scope claims for adversarial follow-up. Answers must preserve the ownership boundary and metric definition in the claim ledger.
Integrity
Never invent implementation details, ownership, benchmark conditions, users, production scale, or business outcomes. Explicitly label suggested future work.