agentleFS
PHY041PHY041/claude-skill-twitter/.claude/skills/twitter-intel/SKILL.md

twitter-intel

Skill5 starsChanged 9 months agoMITView on GitHub

Twitter keyword intelligence — search, monitor, and analyze tweet trends over time. Triggers on 'twitter intel', 'tweet search', 'monitor keyword', 'twitter trend', '/twitter-intel', 'track topic on twitter'.

  • Installs packages

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 1

    1. botlearn-ai/botlearn-skills★ 9same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository2 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 4 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Search Twitter by keyword, collect high-engagement tweets, analyze trends over time, and generate structured reports.

    line 14
  2. Very specific:

    Cookie refresh: When search returns 403, cookies need refresh.

    line 44
  3. Distinctive:

    When user says "search [keyword] on twitter", "twitter intel [topic]", "find tweets about [X]":

    line 50
  4. Distinctive:

    After raw search, filter for quality:

    line 85
  5. Distinctive:

    Read config -> iterate enabled monitors

    line 166
  6. Very specific:

    If any new tweet has favorite_count >= alert_threshold -> immediate alert

    line 171
  7. Distinctive:

    When user says "analyze twitter trend for [keyword]", "twitter trend report":

    line 180

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Twitter Intel — Keyword Search & Trend Monitor

Search Twitter by keyword, collect high-engagement tweets, analyze trends over time, and generate structured reports. Powered by rnet_twitter.py GraphQL search (no browser automation needed).


Architecture

Phase 1: On-demand Search (user-triggered)
  User says "search OpenAI on twitter" -> search -> filter -> report

Phase 2: Keyword Monitoring (cron-driven)
  Config defines keywords -> scheduled search -> diff with last run -> alert on new high-engagement tweets

Phase 3: Trend Analysis (on-demand or weekly)
  Aggregate saved searches -> group by week -> detect topic shifts -> generate narrative

Prerequisites

# Install rnet
pip install "rnet>=3.0.0rc20" --pre

# Required: rnet_twitter.py + cookies
# - rnet_twitter.py (included in this repo, has search_tweets method)
# - twitter_cookies.json (auth_token + ct0, valid ~2 weeks)

Cookie refresh: When search returns 403, cookies need refresh. Get new auth_token + ct0 from Chrome DevTools -> Application -> Cookies -> x.com.


Phase 1: On-demand Search

When user says "search [keyword] on twitter", "twitter intel [topic]", "find tweets about [X]":

Step 1 — Run Search

import asyncio
from rnet_twitter import RnetTwitterClient

async def search(query, count=200):
    client = RnetTwitterClient()
    client.load_cookies("twitter_cookies.json")
    tweets = await client.search_tweets(query, count=count, product="Top")
    return tweets

Search modes:

Modeproduct=Use case
High-engagement"Top"Find influential tweets, content analysis
Real-time"Latest"Monitor breaking discussions, live tracking

Useful Twitter search operators:

OperatorExampleEffect
lang:enOpenAI lang:enEnglish only
since: / until:since:2026-01-24 until:2026-02-24Date range
-filter:repliesOpenAI -filter:repliesOriginal tweets only
min_faves:Nmin_faves:50Minimum likes (only works with Latest)
from:from:karpathySpecific author
"exact""AI agent"Exact phrase

Step 2 — Filter & Enrich

After raw search, filter for quality:

# Filter: relevant + has engagement
filtered = [
    t for t in tweets
    if keyword.lower() in t["text"].lower()  # actually mentions keyword
    and (t["favorite_count"] >= 10 or t["retweet_count"] >= 5)  # has engagement
    and not t["is_reply"]  # original tweets preferred
]

Step 3 — Report

Output a structured summary:

## Twitter Intel: [keyword]
**Period:** [date range] | **Tweets found:** N | **After filter:** N

### Top Tweets (by engagement)
1. @author (X likes, Y RTs, Z views) — date
   "tweet text..."
   [link]

2. ...

### Key Themes
- Theme 1: [description] (N tweets)
- Theme 2: [description] (N tweets)

### Notable Authors
| Author | Followers | Tweets in set | Total engagement |
|--------|-----------|---------------|-----------------|

Phase 2: Keyword Monitoring (Cron)

Config File

{
  "monitors": [
    {
      "id": "my-product-en",
      "query": "MyProduct lang:en -filter:replies",
      "product": "Top",
      "count": 100,
      "min_likes": 10,
      "alert_threshold": 100,
      "enabled": true
    },
    {
      "id": "competitor-mentions",
      "query": "CompetitorName OR \"brand consistency\" lang:en",
      "product": "Latest",
      "count": 50,
      "min_likes": 5,
      "alert_threshold": 50,
      "enabled": true
    }
  ]
}

State File

{
  "my-product-en": {
    "last_run": "2026-02-24T12:00:00Z",
    "last_tweet_ids": ["id1", "id2", "..."],
    "total_collected": 450
  }
}

Cron Workflow

  1. Read config -> iterate enabled monitors
  2. For each monitor:
    • Run search_tweets(query, count, product)
    • Filter by min_likes
    • Diff against last_tweet_ids -> find NEW tweets only
    • If any new tweet has favorite_count >= alert_threshold -> immediate alert
    • Save all new tweets to daily file {monitor_id}/YYYY-MM-DD.json
    • Update state file
  3. Send summary notification (if there are new notable tweets)

Phase 3: Trend Analysis

When user says "analyze twitter trend for [keyword]", "twitter trend report":

Workflow

  1. Load all saved daily files from {monitor_id}/
  2. Group tweets by week
  3. For each week, extract:
    • Total tweet count + total engagement
    • Top 5 tweets by likes
    • Dominant themes (use LLM to categorize)
    • New authors that appeared
    • Sentiment shift
  4. Generate a week-by-week narrative

Output Format

## Trend Report: [keyword]
**Period:** Week 1 (Jan 24-26) to Week 5 (Feb 17-23)
**Total tweets:** N | **Total engagement:** X likes, Y RTs

### Week-by-Week Evolution

#### Week 1 (Jan 24-26): [Theme title]
- Dominant narrative: ...
- Top tweet: @author — "..."
- Key signal: ...

#### Week 2 (Jan 27-Feb 2): [Theme title]
...

### Trend Shifts Detected
1. [Shift description] — happened in Week X
2. ...

### Top Authors Across Period
| Author | Appearances | Total Likes | First seen |

Commands

User SaysAgent Does
/twitter-intel [keyword]Search + filter + report (Top, 200 tweets)
/twitter-intel "[phrase]" --latestSearch Latest mode
monitor "[keyword]" on twitterAdd to monitoring config
twitter intel statusShow all active monitors + last run
twitter trend report [keyword]Analyze saved data, generate trend narrative
refresh twitter cookiesGuide user through cookie refresh

Technical Notes

  • SearchTimeline requires POST (GET returns 404) — this is handled by rnet_twitter.py
  • GraphQL query IDs rotate — if search returns 404, re-extract the SearchTimeline ID from https://abs.twimg.com/responsive-web/client-web/main.*.js
  • User data path (2026-02): screen_name is now at core.user_results.result.core.screen_name (not .legacy)
  • Rate limits: ~300 requests/15min window. With 20 tweets per page, 200 tweets = 10 requests. Safe for cron every 4 hours.
  • Cookie lifetime: auth_token expires after ~2 weeks. Monitor for 403 errors.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.