linkedin-suggest

name	linkedin-suggest
description	Generate a bilingual LinkedIn profile draft (Headline / About / Experience / Skills / Featured / Open to Work) based on ref_src/main.md SSOT. Hybrid mode: one-shot full draft, optional per-field interactive refinement. Use when the user says "/linkedin-suggest", "linkedin draft", "更新 linkedin", "產 linkedin", "linkedin profile", "linkedin 文案", or wants to refresh their LinkedIn copy.

LinkedIn Suggest — Bilingual Profile Draft

Overview

Reads ref_src/main.md (SSOT) and produces ref_src/linkedin_profile.md — a bilingual draft for the 6 main LinkedIn fields. The user copy-pastes each section into LinkedIn UI directly.

Output schema (ref_src/linkedin_profile.md)

# LinkedIn Profile Draft
> Generated by /linkedin-suggest on {date}. Source: ref_src/main.md.

## 1. Headline (EN)
[≤220 chars, punchy, outcome-focused]

## 1. Headline (ZH-TW)
[≤220 chars, same intent in Traditional Chinese]

## 2. About (EN)
[≤2600 chars, narrative first-person, 2-4 short paragraphs]

## 2. About (ZH-TW)
[≤2600 chars]

## 3. Experience — {Company} · {Role} ({Start}–{End}) (EN)
[3-5 bullets, action-verb start, with metrics where possible]

## 3. Experience — {Company} · {Role} (ZH-TW)
[parallel content]

(...repeat per work experience entry...)

## 4. Skills (Top 50, ranked)
1. {skill}
2. {skill}
... (50 lines)

## 5. Featured (suggested links)
- {Project name} → {URL}
- {Project name} → {URL}
- ...

## 6. Open to Work (settings hints)
- Locations: {from main.md Contact section}
- Roles: {derived from Headline + Experience}
- Work types: {Remote / Hybrid / On-site — ask user if unclear}

Workflow

Step 1: Read main.md

Load ref_src/main.md. Extract:

• Name, headline, contact location (from header + Contact section)
• About content (zh-TW + en pairs from ## About)
• All work experience entries (## Work Experience)
• All side projects (## Side Projects)
• Skills + categories (## Core Skills)

If main.md is missing required sections, list what's missing and abort.

Step 2: Field generation (one-shot)

For each of the 6 fields below, follow the corresponding prompt block in PROMPTS.md. Generate English first, then translate/adapt to Traditional Chinese (preserving meaning, not literal). Enforce character limits strictly.

1. Headline
2. About
3. Experience (one block per role from ## Work Experience)
4. Skills (Top 50 ranked by relevance to current target roles)
5. Featured (3-6 best Side Projects with URLs)
6. Open to Work (location + role types)

Step 3: Write ref_src/linkedin_profile.md

Combine all sections into the schema above. Date the header.

If file already exists, ask:

1. Overwrite (default — old version replaced)
2. Save as linkedin_profile-{date}.md (keep history)

Step 4: Print summary

✅ 已產出 ref_src/linkedin_profile.md

字數檢查：
- Headline (EN): {n}/220
- Headline (ZH-TW): {n}/220
- About (EN): {n}/2600
- About (ZH-TW): {n}/2600
- Experience entries: {k}
- Skills: {n}/50
- Featured: {n}/6

要逐欄精修嗎？(y/N)
- 預設 N — 直接複製檔案內容到 LinkedIn 即可
- y — 進入互動模式（每欄印草稿，選 Y/E/R/S）

Step 5 (optional): Interactive refinement

If user picks y, iterate fields 1→6:

For each field, print the current draft, then AskUserQuestion:

Headline (EN) — 220 chars [{current draft}]

1. Y — accept as-is
2. E — provide your edit, I'll save it
3. R — regenerate (give me a hint, e.g. "more emphasis on AI tooling")
4. S — skip (leave as-is, move on)

After loop, rewrite linkedin_profile.md with accepted versions.

Step 6: Done

✅ ref_src/linkedin_profile.md 完成

下一步：
  1. 開檔案，逐區複製
  2. 貼到 LinkedIn 對應欄位（Headline / About / 各 Experience / Skills / Featured / Open to Work）
  3. LinkedIn 對某些欄位有額外格式限制（例如 Experience description 不支援 markdown），複製時用純文字模式

See also

• Field-specific prompts: PROMPTS.md (same dir)
• SSOT: ref_src/main.md