scrape-post

name	scrape-post
description	Given a LinkedIn post URL and a Google Sheet URL, scrape everyone who liked or commented, keep only founders/CEOs/co-founders (ideally SaaS), deduplicate by name, and write Name | Headline | LinkedIn URL | Engagement to the sheet.

LinkedIn Post → Founder List → Google Sheet

Given a LinkedIn post URL and a Google Sheet URL, this skill scrapes every person who liked or commented, filters to founders and CEOs only, deduplicates by name, and writes a clean list to the sheet.

---

Required Inputs

Input	How to extract
Post URL	URL containing linkedin.com/posts/ or linkedin.com/feed/update/
Sheet URL	URL containing docs.google.com/spreadsheets/d/ — extract the Sheet ID from between /d/ and the next /

If the user provides only a post URL, ask for the Google Sheet URL before proceeding.

Strip any tracking params (?utm_source=...) from the post URL before passing to actors.

---

Step 1 & 2: Scrape Reactions and Comments (run in parallel)

Call both actors simultaneously via Apify MCP (call-actor). Use async: false (default) — each call blocks until results are ready.

Reactions — harvestapi/linkedin-post-reactions

{
  "posts": ["<POST_URL>"],
  "maxItems": 1000,
  "profileScraperMode": "short"
}

From each result, extract:

• actor.name → Name
• actor.position → Headline (NOT actor.headline — that field does not exist)
• actor.linkedinUrl → LinkedIn URL
• reactionType → map to engagement label:

Raw value	Label
LIKE	Liked
PRAISE	Celebrated
EMPATHY	Supported
APPRECIATION	Loved
INTEREST	Insightful
ENTERTAINMENT	Funny
(anything else)	(write raw value)

Comments — harvestapi/linkedin-post-comments

{
  "posts": ["<POST_URL>"],
  "maxItems": 1000,
  "profileScraperMode": "short",
  "scrapeReplies": false
}

From each result, extract:

• actor.name → Name
• actor.position → Headline (NOT actor.headline — that field does not exist)
• actor.linkedinUrl → LinkedIn URL
• actor.type → for company filtering
• actor.author → for author filtering
• Engagement label: "Commented"

---

Step 3: Clean and Deduplicate

Apply in this exact order.

3a. Remove company accounts

• Reactions: Drop if actor.linkedinUrl contains /company/
• Comments: Drop if actor.type === "company" or actor.id === null

3b. Remove the post author

• Find any comment record where actor.author === true — store their name (normalized: .trim().toLowerCase())
• Drop that person from both reactions and comments
• If no author found in comments, extract from the post URL slug (e.g. john-doe-12345abc → "john doe") and name-match

3c. Deduplicate by name

Why name-based: Reactions return internal ID URLs (/in/ACoAAA...); comments return vanity slugs (/in/firstname-lastname). Same person → two different URLs. Name dedup catches the overlap.

Dedup key: actor.name.trim().toLowerCase()

When the same person appears in both:

• URL: Keep the vanity slug (the one that does NOT contain ACoAAA). If both are vanity or both are IDs, keep whichever is non-null.
• Headline: Keep whichever actor.position is non-empty
• Engagement: Combine as "<Reaction label> + Commented" (e.g. "Liked + Commented")

If actor.name is blank, fall back to linkedinUrl as dedup key. If both missing, skip the record.

3d. Resolve remaining ACoAAA URLs to vanity slugs

After dedup, check how many records still have internal-ID URLs (contain ACoAAA). If any exist, batch them through harvestapi/linkedin-profile-scraper:

{
  "urls": ["<ACOAAA_URL_1>", "<ACOAAA_URL_2>", "..."],
  "profileScraperMode": "Profile details no email ($4 per 1k)"
}

For each result, replace the ACoAAA URL with the returned linkedinUrl (vanity slug). Use originalQuery.url to match results back to records.

# Build lookup: original ACoAAA url → resolved vanity url
resolved = {item["originalQuery"]["url"]: item["linkedinUrl"] for item in results}
for record in deduplicated:
    if "ACoAAA" in record["linkedin_url"]:
        record["linkedin_url"] = resolved.get(record["linkedin_url"], record["linkedin_url"])

Why after dedup: People who both reacted AND commented already have a vanity slug (from the comments actor). Only pure reactors need resolution — do this after dedup to avoid unnecessary API calls.

Cost: $0.004/profile. Typical post with 80% reactor rate: 68 ACoAAA leads × $0.004 = ~$0.27.

Note: If the actor fails to resolve a URL (network error, profile deleted), keep the ACoAAA URL as-is and flag it — email enrichment will skip it but the lead still appears in the sheet.

---

Step 4: Filter — Founders and CEOs Only

Keep a record if its headline (case-insensitive) contains any of:

founder, co-founder, cofounder, ceo, chief executive, owner, co-ceo

Drop a record if its headline contains any of these (exclusions win over inclusions):

recruiter, talent, hr , investor, vc , venture capital, student, intern

If the headline is blank → drop the record (can't confirm founder status).

After filtering, report: "Filtered N records → M founders kept (X dropped)."

SaaS context: These include keywords are broad enough to catch SaaS founders. No extra SaaS filter is applied — founder/CEO is the primary signal; the headline text itself will often contain "SaaS", "software", "platform", etc. which you can note when relevant but not filter on.

---

Step 5: Write to Google Sheet

Extract the Sheet ID from the user-provided URL (the segment between /d/ and the next /).

First, clear the sheet:

gws sheets spreadsheets values clear \
  --params '{"spreadsheetId": "<SHEET_ID>", "range": "Sheet1"}'

Then write all rows in one call. For any dataset, write JSON to a temp file and use escaped double quotes for --params (single quotes break when --json uses $(...)):

# Build the values array and write to temp file
python3 -c "
import json
rows = [['Name', 'Headline', 'LinkedIn URL', 'Engagement']]
rows += [
  ['Jane Smith', 'Founder & CEO at BuildSaaS', 'https://www.linkedin.com/in/janesmith', 'Liked'],
  # ... all rows
]
print(json.dumps({'values': rows}))
" > /tmp/scrape-post.json

gws sheets spreadsheets values update \
  --params "{\"spreadsheetId\": \"<SHEET_ID>\", \"range\": \"Sheet1!A1\", \"valueInputOption\": \"RAW\"}" \
  --json "$(cat /tmp/scrape-post.json)"

---

Output Columns

Column	Source	Notes
A: Name	actor.name	Full name
B: Headline	actor.position	Role/title (Apify field is position, not headline)
C: LinkedIn URL	actor.linkedinUrl	Prefer vanity slug over internal ID
D: Engagement	Derived	"Liked", "Commented", "Liked + Commented", etc.

---

Key Gotchas (from learnings.md)

• actor.headline does not exist — always use actor.position
• actor.location is not returned in short mode — no location filtering possible without switching to "full" mode (slower, more expensive)
• Reactions URLs are internal IDs (ACoAAA); comments URLs are vanity slugs — deduplicate by name, not URL. After dedup, resolve remaining ACoAAA URLs via harvestapi/linkedin-profile-scraper (Step 3d) — without this, ~80% of reactors can't be email-enriched
• Company accounts: URL contains /company/ (reactions) or actor.type === "company" (comments)
• Set scrapeReplies: false to avoid inflating dataset with author replies
• Always use escaped double-quote --params when --json uses $(cat ...) — single quotes break intermittently