scan-telegram

name	scan-telegram
description	Scans the author's Telegram channel for content suitable for LinkedIn. Analyzes engagement (reactions, views), extracts topics, suggests repackaging. Telegram posts = valid raw source for the DIP and LinkedIn.
disable-model-invocation	true
argument-hint	["Author"]
allowed-tools	Read, Grep, Glob

Scan Telegram for LinkedIn Content

Scan the author's Telegram channel: engagement analysis → topic extraction → repackaging suggestions.

Input

• $0 — author name (Seva, Kirill)
• Optional: path to exported Telegram posts or a folder of files

Required files

1. Author DIP: Knowledge-Base/01-Identity-Profiles/[Author]-DIP.md → Expertise Clusters, Topics, Tone of Voice

2. Algorithm-Intelligence: Knowledge-Base/02-LinkedIn-Algorithm/Algorithm-Intelligence.md → Which formats work on LinkedIn

3. Current-Trends: Knowledge-Base/03-Viral-Trends/Current-Trends.md → What's trending now

4. Telegram source (per author): your-raw-sources/telegram-posts/

5. Posted (to check for duplicates): Posted/[Author]/

Why Telegram = a valuable source

• Telegram posts are written personally by the author (not AI-generated) → valid raw source
• Engagement (reactions, views) shows what resonates with the audience
• The Telegram style is closer to spoken speech → easier to preserve authentic voice
• Topic testing: what landed on Telegram has a high chance of landing on LinkedIn

Pipeline

Step 1: Find Telegram content

1. Identify the author → locate the folder with Telegram posts
2. Read all available posts
3. If there are no posts — tell the user, suggest exporting

Step 2: Engagement Analysis

For each post, extract (if available):

• Views
• Reactions (likes, fire, etc.)
• Reposts
• Comments

Sort by engagement (high to low).

Engagement Score = reactions × 3 + comments × 5 + reposts × 10 (normalized by views)

Step 3: Topic Extraction

For the top 20 posts by engagement:

1. Identify the topic — which Expertise Cluster from the DIP it belongs to
2. Extract the key idea — one sentence
3. Assess LinkedIn potential:
   • Has the topic already appeared on LinkedIn? (check Posted/)
   • Does it fit current trends? (Current-Trends)
   • Is there enough depth for a LinkedIn format?

Step 4: Repackaging Suggestions

For each promising post:

### Telegram → LinkedIn Opportunity [N]

**Telegram post:** [date, first 50 chars]
**Engagement:** [views / reactions / score]
**Topic:** [topic]
**DIP Cluster:** [cluster]
**LinkedIn potential:** HIGH / MEDIUM / LOW

**Why it fits LinkedIn:**
[1–2 sentences]

**Recommended format:**
[from Format-Catalog: story, contrarian take, how-to, etc.]

**Recommended angle:**
[from DIP Viral Angles]

**What the full post needs:**
- [what's already in the Telegram post]
- [what to add: specifics, numbers, case]

**Already on LinkedIn?**
[No / Yes — file, date]

Step 5: Summary Report

## Telegram Scan Report — [Author]

### Scan date: [today]
### Posts analyzed: [N]
### Top topics by engagement:

| # | Topic | Cluster | Engagement Score | LinkedIn potential | Already used? |
|---|-------|---------|------------------|--------------------|---------------|
| 1 | ... | ... | ... | HIGH/MED/LOW | No/Yes |

### Recommendations: what to write next

**Priority 1 (high engagement + not on LinkedIn yet):**
1. [topic] — [why now]

**Priority 2 (medium engagement, but trending):**
1. [topic] — [why now]

**Not recommended (already done or doesn't fit):**
1. [topic] — [why not]

Important

• Telegram posts = INPUT (raw source). They're written personally by the author.
• LinkedIn posts = OUTPUT. Do NOT use LinkedIn posts as a source for new posts.
• When creating a post from Telegram content → cite Telegram as source_ref.
• Recommendations are added to the author's Topic Backlog: Posted/[Author]/Topic-Backlog.md