## Who’s it for Growth hackers, community builders, and marketers who want to keep their Twitter (X) accounts active by *liking posts from selected profiles* automatically. ## How it works / What it does 1. **Sche
The AI agent pulls profile URLs from a CSV, schedules hourly actions, and fetches recent posts. It rotates session cookies to simulate varied logins, selects posts that haven't been liked yet, and compiles a list to process. It performs the likes via Phantombuster, then logs each liked URL and uploads the results to SharePoint for audit.
Automates engagement by querying a profile list, rotating credentials, and recording outcomes.
Fetches up to 20 tweets per profile from the provided CSV.
Rotates Twitter session cookies to simulate different logins.
Checks against the already-liked log to avoid duplicates.
Creates twitter_posts_to_like.csv and uploads it to SharePoint.
Executes likes via Phantombuster Autolike Agent.
Logs each liked URL to prevent duplicates.
This AI agent replaces fragmented, manual engagement with a repeatable, auditable process. It standardizes hourly activity across many profiles while keeping a clear log for reporting.
A simple 3-step AI agent flow.
Imports profile URLs from the profiles CSV and schedules hourly runs.
Rotates Twitter cookies, fetches up to 20 tweets per profile, and filters out already liked posts.
Likes eligible posts via Phantombuster and logs the URLs to SharePoint.
One realistic scenario demonstrates how the AI agent handles a campaign.
Scenario: A Growth Marketer uses a profiles_twitter.csv with five profiles and a twitter_session_cookies.txt file. The AI agent runs hourly for 6 hours, processes up to 20 tweets per profile, and likes 60–100 new posts. It uploads twitter_posts_to_like.csv to SharePoint and updates twitter_posts_already_liked.csv for auditing.
One supporting sentence.
needs scalable, profile-targeted engagement without manual checks.
maintains active presence across multiple communities.
delivers consistent engagement for multiple clients.
requires auditable engagement logs for reporting.
initiates engagement on multiple profiles while tracking results.
scales outreach across profiles without manual workload.
Phantombuster and SharePoint are used to run and store results.
Executes the actual like actions on Twitter for the selected posts.
Stores the twitter_posts_to_like.csv, logs, and performs auditing.
Provides the target profile URLs that drive the engagement workflow.
Six practical scenarios to apply this AI agent.
One supporting sentence with practical concerns.
The AI agent operates within commonly accepted engagement practices and relies on user-configured parameters to avoid spam-like behavior. It uses documented workflows and provides explicit opt-ins for any mass engagement. However, platform policies can change, so it’s important to review current rules before deployment. If you plan to run this in production, configure frequency and profile lists conservatively and monitor outcomes.
Yes. The AI agent is designed to run on an hourly schedule by default, but you can adjust timing in the configuration. You can also set limits on posts per profile and the total number of profiles processed in a run. Changes apply to subsequent runs and are auditable in the SharePoint log.
Duplicates are prevented by maintaining a log of already-liked posts in a CSV and by checking new posts against that log before execution. The system updates twitter_posts_already_liked.csv after each run. If a post previously liked is encountered again, it is skipped automatically.
All actions are logged and stored in SharePoint. The log includes the URLs of liked posts, timestamps, and the profiles involved. This creates an auditable trail suitable for reporting or client reviews. Access controls can be applied to protect sensitive data.
If a profile URL changes, update the profiles CSV and re-run. The AI agent reads the latest CSV on each run, so changes are incorporated automatically. If a profile is temporarily unavailable, it will be skipped and retried in the next cycle. You’ll still retain a log of attempts for visibility.
SharePoint is used to store the output CSVs and logs, but the core logic can be adapted to alternative storage. If SharePoint is unavailable, you can configure the agent to save outputs to a local file or another system, though auditing and collaboration features may be reduced. Any changes should be tested to ensure data integrity.
To stop the AI agent, disable the scheduled task or pause the configuration in the control panel. Ensure any in-progress runs are finished gracefully to avoid partial writes. When re-enabled, the agent will resume from the next scheduled hour. You can also export current settings for backup before stopping.
## Who’s it for Growth hackers, community builders, and marketers who want to keep their Twitter (X) accounts active by *liking posts from selected profiles* automatically. ## How it works / What it does 1. **Sche
