anuchit1tube168-cmd

RTAFNC Advisor System Skill

- no real student data is committed to public GitHub

anuchit1tube168-cmd 0 Updated 1mo ago
GitHub

Install

npx skillscat add anuchit1tube168-cmd/gooddeeds69/skills-rtafnc-advisor

Install via the SkillsCat registry.

SKILL.md

RTAFNC Advisor System Skill

Purpose

ใช้สกิลนี้เมื่อออกแบบ/พัฒนา/ตรวจระบบอาจารย์ที่ปรึกษา วิทยาลัยพยาบาลทหารอากาศ (RTAFNC ONE) โดยยึดคู่มืออาจารย์ที่ปรึกษา ปีการศึกษา 2566 เป็น baseline และไม่ทำลายระบบ Production เดิม

Core principle

One student identity → one advisor assignment history → one consultation timeline → one annual development summary.

Source-backed backbone

  • อษ.1 ทะเบียนประวัตินักเรียนพยาบาล: ข้อมูลพื้นฐานและครอบครัว ใช้เพื่อรู้จักผู้เรียนและอัปเดตตามระยะ
  • อษ.2 ระเบียนสะสม: ผลการศึกษา ทุน/รางวัล กิจกรรม บทบาท และพัฒนาการตามปีการศึกษา
  • อษ.3 บันทึกการพบ: บันทึกรายบุคคล ทั้งการพบเป็นทางการ/ไม่เป็นทางการหรือเหตุที่ควรบันทึก และควรมีอย่างน้อยเดือนละ 1 ครั้ง
  • อษ.4 สรุปผลพัฒนาการ: สรุปภาพรวมอย่างน้อยปีละ 1 ครั้ง

Primary consultation domains

คู่มือกำหนดหัวข้อหลัก 4 กลุ่ม:

  1. การศึกษา/การเรียนการสอน
  2. เรื่องส่วนตัว
  3. เรื่องสังคม
  4. งานอาชีพ/อนาคต

Operational subcategories ที่ระบบใช้เพื่อค้นหา/รายงานได้ละเอียดขึ้น โดย map กลับเข้าหมวดหลักเสมอ:

  • Academic / Learning
  • Personal / Family
  • Social / Peer / Adjustment
  • Career / Professional goal
  • Health / Physical
  • Emotional / Mental well-being
  • Financial / Scholarship
  • Discipline / Conduct
  • Military/social adjustment
  • Activities / Leadership
  • Other

Standard workflow

  1. Student opens Advisor LIFF.
  2. Resolve RTAFNC_ID and active advisor assignment.
  3. Student requests meeting or advisor creates meeting record.
  4. Select purpose/topic, urgency and preferred channel/time.
  5. Advisor receives Telegram primary alert with secure link.
  6. Advisor accepts/reschedules/refers.
  7. Meeting occurs: onsite / phone / secure chat / video / other approved channel.
  8. Advisor completes digital อษ.3:
    • date/time
    • formal/informal
    • topic/domain
    • concise facts / student concern
    • advisor observation
    • guidance/counseling provided
    • what student gained/understood
    • agreed action plan
    • follow-up date
    • referral if needed
  9. Student may receive a non-sensitive meeting summary and next action.
  10. Follow-up task closes or escalates.
  11. At academic-year summary, generate digital อษ.4 from verified data + advisor narrative.
  12. Export authorized reports to PDF/Word and archive privately in Drive.

Counseling method checklist

During a consultation, guide the advisor through:

  • Build rapport / safe atmosphere
  • Listening and observation
  • Leading / clarification
  • Reflection
  • Summarization
  • Informing
  • Encouragement
  • Suggestion while preserving student decision-making
  • Referral when beyond advisor scope

Minimum record rules

  • Do not force advisor to write detailed sensitive notes while the student is watching.
  • อษ.3 target: at least 1 individual record/student/month when applicable per handbook baseline.
  • อษ.4 target: at least 1 annual summary/student/academic year.
  • Every record has academic_year, student_id, advisor_id, timestamp, source, version, status and audit metadata.

Confidentiality & access

  • Consultation notes are confidential by default.
  • Student sees only student-safe summary, not internal advisor notes unless policy explicitly permits.
  • Advisor sees assigned students within assignment dates.
  • Head/governance roles see only authorized scope.
  • General admin/IT role does NOT automatically grant access to sensitive counseling notes.
  • Reports containing confidential information should prefer student code over full name unless the recipient is authorized and a name is required.
  • Never put detailed health/mental-health/disciplinary content in Telegram or LINE lock-screen notifications; send case ID + urgency + secure link.
  • All access/download/export actions must be audit logged.

Assignment model

AdvisorAssignment:

  • assignment_id
  • student_id
  • advisor_id
  • academic_year
  • start_at
  • end_at
  • status
  • assigned_by
  • note

Support reassignment without overwriting history.
Baseline planning ratio from handbook: advisor should normally have no more than 1:8 students where applicable.

Data model

StudentProfile (prefill from RTAFNC Master)
AdvisorAssignment
ConsultationRequest
ConsultationSession
ConsultationTopic
ConsultationOutcome
FollowUpTask
Referral
AdvisorNotePrivate
StudentSummarySafe
AnnualDevelopmentSummary (อษ.4)
AdvisorAvailability
NotificationEvent
ConsentPrivacyReceipt
AuditLog

Prefill sources

Reuse verified existing data instead of retyping:

  • student name / code / cohort / year / photo
  • advisor assignment
  • GPAX / academic progress
  • scholarship history
  • good-deed / volunteer-hour summaries
  • health case status only where purpose/role permits; do not expose clinical detail by default
  • activities / awards / leadership roles
  • conduct data only for authorized workflow

Reporting requirements

Filters:

  • academic year
  • semester
  • advisor
  • cohort/year
  • student
  • topic/domain
  • session status
  • referral/follow-up status

Core metrics:

  • number of assigned students per advisor
  • number of students who met advisor
  • total consultations
  • consultations per student
  • consultations per advisor
  • monthly trend
  • topic distribution
  • formal vs informal encounters
  • follow-ups due/completed/overdue
  • referrals by destination category
  • annual อษ.4 completion rate

Do not rank advisors by raw consultation counts alone. Counts are operational evidence, not a quality score.

Outputs

  1. Student LIFF: advisor card, request appointment, history-safe summary, next actions.
  2. Advisor LIFF: group dashboard, student list, pending requests, meeting form, follow-up queue, annual summary.
  3. Admin: assignment management, academic year rollover, report center, template/version control.
  4. Telegram: primary operational notification + secure deep link.
  5. LINE: urgent/special notification only according to RTAFNC notification policy.
  6. Word/PDF: digital อษ.1–อษ.4 and management reports.

Link pattern

Never place sensitive data in query strings.
Use opaque short-lived tokens, e.g.
/advisor.html?task=ADV-<opaque-token>
Backend resolves task → verifies identity/role/purpose → returns authorized data.

Year rollover

At start of academic year:

  • create AcademicYear config
  • roll student year level
  • keep prior advisor assignment history
  • allow admin to continue/change assignments
  • reset annual report counters while preserving historical sessions
  • create annual อษ.4 task per active student

Definition of done

A module is not complete until:

  • student can request/see safe status
  • advisor can receive link and record session
  • follow-up/referral works
  • admin can assign/reassign advisor
  • yearly/monthly report works
  • Word/PDF export works
  • access controls and audit are tested
  • sensitive notification redaction is tested
  • no real student data is committed to public GitHub