JevMade field notes / Architecture and workflow guide

A guide to running social media research with an AI tool

This guide explains how to set up a tool that lets an AI choose specific actions to search social media. You will learn how it gathers real evidence and handles roadblocks like login screens or errors.

Original by socai-ioBrowser use

Listen to this guide

JevMade’s plain-English explanation

0:00 /

AI narration

Credits and license

“Jev Social — browser-grounded social research” by socai-io. Read the original source.

This expanded guide is an AI-narrated adaptation of the source’s essential explanation, examples and caveats, not a word-for-word reading. The synthetic voice does not imitate the author or imply their endorsement.

License: MIT

MIT License

Copyright (c) 2026 jev-social contributors

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

Our summary

Jev Social is a project that helps people gather information from platforms like Instagram, TikTok, and LinkedIn. It uses an AI tool named Jev that chooses from a list of specific options rather than writing its own answers. This helps researchers collect real, linked evidence instead of guessed information.

The software works by giving the AI a strict list of allowed actions, like searching or reading comments. When the AI chooses an action, a separate browser tool carries it out on a real web page. The system checks every result to make sure the final report only includes actual findings.

This setup is useful for people who need reliable records of social media posts. However, it has strict limits. If a website asks for a login or the software runs out of steps, it stops and saves what it has found so far rather than claiming it finished the whole task.

Key takeaways

The AI can only choose from a strict list of allowed actions and previously found links.
If the software hits a login screen or an error, it saves the partial information it found.
The creators require at least ten separate test runs before they publish any performance summaries.

More guides →See the experiments →