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

twitter-cultivate

Skill5 starsChanged 9 months agoMITView on GitHub

Twitter/X account cultivation and growth system. Checks account health (TweepCred, shadowban), analyzes tweets, finds engagement opportunities, recommends unfollows, and tracks progress. Triggers on "/twitter-cultivate", "check my twitter", "twitter health", "grow my twitter", "twitter maintenance", "fix my twitter reach".

  • Installs packages

1 other project uses a version of this, 1 of them edited.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 1

    1. PHY041/claude-agent-skills★ 1875% same

    Wrote their own version · 0

    None found.

    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

11 rules. About 0 are common, 10 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈10

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    No tweets in 90+ days (inactive)

    line 97
  2. Distinctive:

    Never interacted with you (no value)

    line 98
  3. Very specific:

    Use search_tweets to find trending conversations:

    line 104
  4. Distinctive:

    No tweets in 90+ days

    line 154
  5. Distinctive:

    Never liked/replied to your tweets

    line 158
  6. Distinctive:

    No content in your niche

    line 163
  7. Distinctive:

    Every tweet should have ONE of:

    line 200
  8. Distinctive:

    If your ratio is > 5.0 (following >> followers):

    line 305
  9. Distinctive:

    If you have 0% media tweets:

    line 311
  10. Distinctive:

    Add image/video to EVERY tweet

    line 312
  11. Distinctive:

    Use Canva/Figma for quick graphics

    line 313

≈ “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 Account Cultivation Skill

Systematic approach to growing Twitter presence based on the open-source algorithm analysis.


Prerequisites

  • rnet installed (pip install "rnet>=3.0.0rc20" --pre)
  • rnet_twitter.py — the lightweight GraphQL client included in this repo
  • Twitter cookies exported to: twitter_cookies.json Format: [{"name": "auth_token", "value": "..."}, {"name": "ct0", "value": "..."}]
  • Your Twitter handle configured

Getting Cookies

  1. Open Chrome → go to x.com → log in
  2. DevTools → Application → Cookies → https://x.com
  3. Copy auth_token and ct0 values
  4. Save to twitter_cookies.json (see twitter_cookies.example.json)
  5. Cookies last ~2 weeks. Refresh when you get 403 errors.

Core Metrics to Track

MetricHealthy RangeImpact
Following/Follower Ratio< 0.6TweepCred score
Avg Views/Tweet20-40% of followersAlgorithm favor
Media Tweet %> 50%10x engagement
Link Tweet %< 20%Avoid algorithm penalty
Reply RateReply to 100% of comments+75 weight boost

Workflow: Full Health Check

Step 1: Analyze Account

import asyncio
from rnet_twitter import RnetTwitterClient

async def analyze(username: str):
    client = RnetTwitterClient()
    client.load_cookies("twitter_cookies.json")

    # Get user profile
    user = await client.get_user_by_screen_name(username)
    followers = user.get("followers_count", 0)
    following = user.get("friends_count", 0)
    ratio = following / max(followers, 1)

    # Get recent tweets for content analysis
    tweets = await client.get_user_tweets(user["rest_id"], count=20)

    return {
        "username": username,
        "followers": followers,
        "following": following,
        "ratio": round(ratio, 2),
        "tweet_count": user.get("statuses_count", 0),
        "recent_tweets": len(tweets),
    }

asyncio.run(analyze("YOUR_USERNAME"))

Step 2: Check Shadowban Status

Manual check: shadowban.yuzurisa.com

Step 3: Analyze Following List

Recommends accounts to unfollow based on:

  • No tweets in 90+ days (inactive)
  • Never interacted with you (no value)
  • Low follower count + high following (likely bots)
  • No mutual engagement

Step 4: Find Engagement Opportunities

Use search_tweets to find trending conversations:

async def find_opportunities(niche_keywords: list[str]):
    client = RnetTwitterClient()
    client.load_cookies("twitter_cookies.json")

    opportunities = []
    for keyword in niche_keywords:
        tweets = await client.search_tweets(
            f"{keyword} lang:en -filter:replies",
            count=50,
            product="Top"
        )
        # Filter for high-engagement, recent tweets
        for t in tweets:
            if t["favorite_count"] >= 50 and t["reply_count"] < 20:
                opportunities.append(t)

    return sorted(opportunities, key=lambda t: t["favorite_count"], reverse=True)

Step 5: Generate Weekly Report

Compile metrics from Steps 1-4 into a structured report.


Account Health Scoring

Based on Twitter's open-source algorithm:

TweepCred Estimation

Score = PageRank x (1 / max(1, following/followers))
RatioEstimated TweepCredAlgorithm Treatment
< 0.665+ (healthy)All tweets considered
0.6 - 2.040-65Limited consideration
2.0 - 5.020-40Severe penalty
> 5.0< 20Only 3 tweets max

Unfollow Strategy

Priority 1: Inactive Accounts

  • No tweets in 90+ days
  • Safe to unfollow, no relationship loss

Priority 2: Non-Engagers

  • Never liked/replied to your tweets
  • One-way relationship

Priority 3: Low-Value Follows

  • High following/low followers (bot-like)
  • No content in your niche

Execution Plan

Week 1: Unfollow 30 inactive accounts
Week 2: Unfollow 30 non-engagers
Week 3: Unfollow 30 low-value follows
Week 4: Evaluate ratio improvement

Target: Get ratio below 2.0, ideally below 0.6


Content Strategy (Algorithm-Optimized)

Tweet Types by Algorithm Weight

TypeWeightRecommendation
Tweet that gets author reply+75ALWAYS reply to comments
Tweet with replies+13.5Ask questions
Tweet with profile clicks+12.0Be intriguing
Tweet with long dwell time+10.0Use threads
Retweet+1.0Low value
Like+0.5Lowest value

Content Mix

  • 40% Value content (insights, tips, frameworks)
  • 30% Engagement bait (questions, polls, hot takes)
  • 20% Build-in-public (progress updates, wins, losses)
  • 10% Promotion (with value attached)

Media Requirements

Every tweet should have ONE of:

  • Image (infographic, screenshot, meme)
  • Video (< 2:20, hook in first 3 sec)
  • Poll
  • Thread (7-10 tweets)

NEVER post text-only tweets


Posting Schedule

Optimal Times (General)

DayBest TimeSecond Best
Tuesday9-10 AM1-2 PM
Wednesday9-10 AM3-4 PM
Thursday10-11 AM2-3 PM

First 10 Minutes Protocol

1. Post at optimal time
2. Immediately self-reply with additional insight
3. Reply to ANY comment within 10 minutes
4. Have 2-3 "pod" members ready to RT

Frequency

  • Minimum: 1 tweet/day
  • Optimal: 3-5 tweets/day
  • Gap: 30-60 min between tweets

Engagement Tactics

Reply Strategy (Most Important)

The algorithm gives +75 weight when you reply to replies on your tweets.

Someone comments on your tweet
    |
Reply within 30 minutes (CRITICAL)
    |
Algorithm sees author engagement
    |
Tweet gets boosted to more feeds

Quote Tweet Strategy

Find viral tweet in your niche
    |
Quote with your unique take
    |
Add value, not just "great point"
    |
Post during optimal hours

Thread Formula

1/ Hook (curiosity gap or bold claim)
2-6/ Individual points with specifics
7/ Summary
8/ CTA: Question or "follow for more"

Weekly Routine

Daily (15 min)

  • Post 1-3 tweets with media
  • Reply to ALL comments on your tweets
  • Engage with 5-10 tweets in your niche
  • Check notifications and respond

Weekly (Saturday)

  • Run full health check
  • Review what content performed best
  • Unfollow 10-20 low-value accounts
  • Plan next week's content themes

Monthly

  • Full ratio review (target < 2.0)
  • Shadowban check
  • Content audit (media %, link %)
  • Milestone check (follower goals)

Recovery Plan (Low Follower Count)

Phase 1: Emergency Ratio Fix (Week 1-2)

If your ratio is > 5.0 (following >> followers):

  • Unfollow 100+ inactive/non-engaging accounts
  • Target: ratio < 5.0 as first milestone

Phase 2: Content Upgrade (Week 2-4)

If you have 0% media tweets:

  • Add image/video to EVERY tweet
  • Use Canva/Figma for quick graphics
  • Screenshot interesting data/insights

Phase 3: Engagement Building (Week 3-6)

  • Reply to 20+ tweets/day in your niche
  • Quote tweet viral content with your take
  • Join relevant Twitter communities
  • DM potential collaborators

Phase 4: Consistency (Ongoing)

  • 3-5 tweets/day
  • Reply to 100% of comments
  • Weekly analysis and adjustment

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.