How do I install the Paid Collab Extractor skill?

Run npx skillmds@latest add shannhk/paid-collab-extractor in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Paid Collab Extractor safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Paid Collab Extractor?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Paid Collab Extractor free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Paid Collab Extractor?

shannhk (@shannhk) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Paid Collab Extractor by shannhk — Agent Skill

Name: Paid Collab Extractor
Author: shannhk

Paid Collab Extractor

Extract every creator paid for a collab on an X/Twitter post via paid.airaa.xyz. Use when the user gives a tweet/X URL and asks to find paid creators, paid collaborators, sponsored creators, who was paid to promote a tweet, who got paid for the collab, or references paid.airaa.xyz, airaa.xyz, or Airaa. Outputs a CSV with handle, display name, followers, tier (Nano/Mid/Macro/Mega), likes, RTs, views, snippet, quote-tweet URL, and profile URL. Triggers on "who was paid for this tweet", "paid collabs", "airaa", "paid promotions", "who got paid to post this", "sponsored creators on this tweet", "extract paid collaborators", "/paid-collab-extractor".

shannhk Updated Sep 17, 2026

File contents

Paid Collab Extractor

Drives headless Chromium via Playwright, feeds an X post URL to paid.airaa.xyz, waits for the scan to finish, and writes a CSV of every paid creator listed.

No LLM calls, no API keys — pure deterministic Playwright + BeautifulSoup.

Steps

Confirm the input URL. Expect an X post URL like https://x.com/user/status/1234.... If the user didn't give one, ask for one before continuing.

Bootstrap the venv (first run only). Check for the skill's .venv:

cd ~/.claude/skills/paid-collab-extractor && \
  [ -d .venv ] || ( \
    python3 -m venv .venv && \
    source .venv/bin/activate && \
    pip install --quiet -r scripts/requirements.txt && \
    python -m playwright install chromium \
  )

This takes ~30–60s the first time (downloads Chromium). Subsequent runs skip it.

Run the extractor. Default output is airaa_results.csv. If the tweet URL contains a status ID, prefer <handle>_<status_id>.csv for clarity.
```
source ~/.claude/skills/paid-collab-extractor/.venv/bin/activate && \
  python ~/.claude/skills/paid-collab-extractor/scripts/extract.py \
    "<TWEET_URL>" --out "<OUTPUT_PATH>"
```
Output format: auto-detected from the --out extension.
- .csv (default) → flat CSV
- .json → structured JSON with tweet_url, source_handle, tweet_id, scanned_at, count, and creators[]. Use JSON when feeding data into a downstream app (e.g. the paid-collab dashboard's /data directory).
Force a format with --format csv|json if the extension is ambiguous.
Report results. Show the user:
- Total paid creators found
- Absolute path to the CSV
- A 5-row preview sorted by followers (the site's default order)

Output schema

Columns produced in the CSV:

Column	Example	Notes
`handle`	`@historyinmemes`	X handle with `@` prefix
`name`	`Historic Vids`	Display name pulled from avatar `alt`
`avatar_url`	`https://pbs.twimg.com/profile_images/.../X_bigger.jpg`	Creator's profile picture (JSON only)
`followers`	`6.2M`	Raw string as rendered (K/M suffix)
`tier`	`Mega`	Nano / Mid / Macro / Mega
`likes`	`45`	On the creator's quote-tweet
`rts`	`3`	Retweets on the QT
`views`	`60.3K`	Views on the QT
`snippet`	`In the Medieval Era...`	First line of the QT text
`qt_url`	`https://x.com/.../status/...`	That creator's QT
`profile_url`	`https://x.com/historyinmemes`	Creator's X profile

JSON files also include top-level metadata: tweet_url, source_handle, source_name (the advertiser's display name, fetched via X oEmbed — e.g. "Higgsfield AI" for @higgsfield), tweet_id, scanned_at, count.

Edge cases

No paid collabs. The scan completes with 0 paid found. Report that plainly instead of producing an empty CSV.
Scan stalls. Max wait is 120s. If counts stop moving for ~9s the script exits with whatever it has. This is normal even for complete scans.
Private/deleted tweet. The site may show an error state. Relay the error text to the user rather than pretending the scan worked.
Zero results repeatedly. Back off a few minutes before retrying — likely transient rate limiting.

Chaining

After running, common next steps:

Filter to Mega/Macro tier with awk -F, '$4 ~ /Mega|Macro/' airaa_results.csv
Pull each creator's full profile → chain to twitter-scrape-analyze
Draft outreach → chain to direct-response-copy

---name: paid-collab-extractordescription: Extract every creator paid for a collab on an X/Twitter post via paid.airaa.xyz. Use when the user gives a tweet/X URL and asks to find paid creators, paid collaborators, sponsored creators, who was paid to promote a tweet, who got paid for the collab, or references paid.airaa.xyz, airaa.xyz, or Airaa. Outputs a CSV with handle, display name, followers, tier (Nano/Mid/Macro/Mega), likes, RTs, views, snippet, quote-tweet URL, and profile URL. Triggers on "who was paid for this tweet", "paid collabs", "airaa", "paid promotions", "who got paid to post this", "sponsored creators on this tweet", "extract paid collaborators", "/paid-collab-extractor".---# Paid Collab ExtractorDrives headless Chromium via Playwright, feeds an X post URL to `paid.airaa.xyz`, waitsfor the scan to finish, and writes a CSV of every paid creator listed.No LLM calls, no API keys — pure deterministic Playwright + BeautifulSoup.## Steps1. **Confirm the input URL.** Expect an X post URL like `https://x.com/user/status/1234...`. If the user didn't give one, ask for one before continuing.2. **Bootstrap the venv (first run only).** Check for the skill's `.venv`: ```bash cd ~/.claude/skills/paid-collab-extractor && \ [ -d .venv ] || ( \ python3 -m venv .venv && \ source .venv/bin/activate && \ pip install --quiet -r scripts/requirements.txt && \ python -m playwright install chromium \ ) ``` This takes ~30–60s the first time (downloads Chromium). Subsequent runs skip it.3. **Run the extractor.** Default output is `airaa_results.csv`. If the tweet URL contains a status ID, prefer `<handle>_<status_id>.csv` for clarity. ```bash source ~/.claude/skills/paid-collab-extractor/.venv/bin/activate && \ python ~/.claude/skills/paid-collab-extractor/scripts/extract.py \ "<TWEET_URL>" --out "<OUTPUT_PATH>" ``` **Output format:** auto-detected from the `--out` extension. - `.csv` (default) → flat CSV - `.json` → structured JSON with `tweet_url`, `source_handle`, `tweet_id`, `scanned_at`, `count`, and `creators[]`. Use JSON when feeding data into a downstream app (e.g. the paid-collab dashboard's `/data` directory). Force a format with `--format csv|json` if the extension is ambiguous.4. **Report results.** Show the user: - Total paid creators found - Absolute path to the CSV - A 5-row preview sorted by followers (the site's default order)## Output schemaColumns produced in the CSV:| Column | Example | Notes ||---|---|---|| `handle` | `@historyinmemes` | X handle with `@` prefix || `name` | `Historic Vids` | Display name pulled from avatar `alt` || `avatar_url` | `https://pbs.twimg.com/profile_images/.../X_bigger.jpg` | Creator's profile picture (JSON only) || `followers` | `6.2M` | Raw string as rendered (K/M suffix) || `tier` | `Mega` | Nano / Mid / Macro / Mega || `likes` | `45` | On the creator's quote-tweet || `rts` | `3` | Retweets on the QT || `views` | `60.3K` | Views on the QT || `snippet` | `In the Medieval Era...` | First line of the QT text || `qt_url` | `https://x.com/.../status/...` | That creator's QT || `profile_url` | `https://x.com/historyinmemes` | Creator's X profile |JSON files also include top-level metadata: `tweet_url`, `source_handle`, `source_name` (the advertiser's display name, fetched via X oEmbed — e.g. `"Higgsfield AI"` for `@higgsfield`), `tweet_id`, `scanned_at`, `count`.## Edge cases- **No paid collabs.** The scan completes with `0 paid found`. Report that plainly instead of producing an empty CSV.- **Scan stalls.** Max wait is 120s. If counts stop moving for ~9s the script exits with whatever it has. This is normal even for complete scans.- **Private/deleted tweet.** The site may show an error state. Relay the error text to the user rather than pretending the scan worked.- **Zero results repeatedly.** Back off a few minutes before retrying — likely transient rate limiting.## ChainingAfter running, common next steps:- Filter to Mega/Macro tier with `awk -F, '$4 ~ /Mega|Macro/' airaa_results.csv`- Pull each creator's full profile → chain to `twitter-scrape-analyze`- Draft outreach → chain to `direct-response-copy`

Frequently asked questions

Paid Collab Extractor

File contents

Paid Collab Extractor

Steps

Output schema

Edge cases

Chaining

Related skills

Frequently asked questions

How do I install the Paid Collab Extractor skill?

What does the Paid Collab Extractor skill do?

Is Paid Collab Extractor safe to use?

Which AI agents work with Paid Collab Extractor?

Is Paid Collab Extractor free to use?

Who published Paid Collab Extractor?