Build Interview Q&A

interview-qa sources candidate questions and drafts first-person answers for the skill sheet's Q&A (FAQ) section.

Both question and answer are published as your own words, so the AI only proposes and drafts; it saves only after explicit approval.

What you can do

Ask what you've actually been asked in real interviews (the highest-confidence source)
Derive candidate questions from category distribution and memo clusters (public and private)
Research common technical interview questions on the web and filter them to your skills and domains (falls back to general model knowledge, clearly labeled as tentative, if web search isn't available)
Draft a first-person answer for each adopted question, grounded in related memos when they exist
Save the approved Q&A to your skill sheet
Propose saved answers as reusable knowledge candidates through the paput-capture workflow when they qualify

Skill to use

In Claude, call this skill:

/paput-interview-qa

Via the plugin, call /paput:interview-qa. In clients such as Codex where slash-command skills are unavailable, ask in natural language to use interview-qa.

Main MCP tools used internally

The skill normally guides the required tool calls. Internally, it mainly uses these tools.

Use paput_get_skill_sheet to fetch the current Q&A, skills, domains, and projects
Use paput_get_categories and paput_search_memo to find memo clusters and derive candidate questions
Draft an answer for each adopted question in the AI client
Present the candidates and drafts, and confirm which to adopt, reject, or save
After explicit approval, re-fetch the latest Q&A with paput_get_skill_sheet, merge in the changes, and save with paput_update_skill_sheet_faq
Report any dropped_ids returned when related memo IDs are dropped on save

Notes

It never auto-saves. Adopting questions and saving answers both require explicit user approval
paput_update_skill_sheet_faq is a full replace. Each save resends the complete Q&A list (an empty array clears everything)
Linking related memos (related_memo_ids) is optional. Only public memos you own are resolved; anything else is silently dropped on save
You can register up to 15 Q&A items. Aim for 10-15 and prioritize rather than keeping everything
No canonical list of interview questions ships with the skill. Interview trends change, so it researches the web each time instead

What you can do

Ask what you've actually been asked in real interviews (the highest-confidence source)

Derive candidate questions from category distribution and memo clusters (public and private)

Research common technical interview questions on the web and filter them to your skills and domains (falls back to general model knowledge, clearly labeled as tentative, if web search isn't available)

Draft a first-person answer for each adopted question, grounded in related memos when they exist

Save the approved Q&A to your skill sheet

Propose saved answers as reusable knowledge candidates through the paput-capture workflow when they qualify

Main MCP tools used internally

The skill normally guides the required tool calls. Internally, it mainly uses these tools.

Use paput_get_skill_sheet to fetch the current Q&A, skills, domains, and projects

Use paput_get_categories and paput_search_memo to find memo clusters and derive candidate questions

Draft an answer for each adopted question in the AI client

Present the candidates and drafts, and confirm which to adopt, reject, or save

After explicit approval, re-fetch the latest Q&A with paput_get_skill_sheet, merge in the changes, and save with paput_update_skill_sheet_faq

Report any dropped_ids returned when related memo IDs are dropped on save

Notes

It never auto-saves. Adopting questions and saving answers both require explicit user approval

paput_update_skill_sheet_faq is a full replace. Each save resends the complete Q&A list (an empty array clears everything)

Linking related memos (related_memo_ids) is optional. Only public memos you own are resolved; anything else is silently dropped on save

You can register up to 15 Q&A items. Aim for 10-15 and prioritize rather than keeping everything

No canonical list of interview questions ships with the skill. Interview trends change, so it researches the web each time instead