AgentSkillsCN

kb-answer-questions

批量解决待处理问题或缺失信息:运行 python3 scripts/kb.py questions(可选指定日期),请用户确认时间追踪占位符、待审核事项、Jira 清理工作,以及手动提出的开放性问题;随后,通过更新 notes/work/end-of-day.md 和 notes/work/open-questions.md 来应用解答。可通过 /answer-questions、“我可以回答问题”、“梳理问题”等指令触发。

SKILL.md
--- frontmatter
name: kb-answer-questions
description: 'Batch-resolve pending questions/missing info: run python3 scripts/kb.py questions (optionally for a date), ask the user to confirm time tracking placeholders, needs-review items, Jira hygiene, and manual open questions; then apply answers by updating notes/work/end-of-day.md and notes/work/open-questions.md. Triggered by /answer-questions, "могу ответить на вопросы", "разобрать вопросы".'

kb-answer-questions

Use this skill when the user is ready to answer questions / resolve missing info.

Steps

  1. Anchor time:
bash
date '+%Y-%m-%d %H:%M (%A) %Z'
  1. Generate the pending questions report:
bash
.venv/bin/python scripts/kb.py questions

If the user specified a target day (YYYY-MM-DD):

bash
.venv/bin/python scripts/kb.py questions --date YYYY-MM-DD
  1. Ask the user to resolve items in batch (minimal interaction):
  • Time tracking: ask for hours only where __h is present.
  • Needs review: ask what to do (confirm theme mapping / ignore for now).
  • Jira hygiene: ask whether to update status/comment (yes/no) for listed keys.
  • Manual open questions: answer/close them or keep open.
  1. Apply answers:
  • Update notes/work/end-of-day.md → “Time Tracking (draft)” with confirmed hours (do not invent).
  • Move resolved items from notes/work/open-questions.md → “Resolved (log)” with a short trail.

Output expectations

  • Return a short checklist: what was resolved vs what remains open.