Skip to main content
Once your workspace is connected, Neo can be used directly inside Slack for fast, collaborative security work.

​
Where You Can Use Neo

Neo works in two main modes:
  • Direct messages for focused one-on-one interaction with Neo
  • Channels and threads for team-visible investigations and follow-up
In channels, threading matters. Neo keeps work scoped to a thread so the conversation, task status, and final summary stay in one place.

​
DMs vs Channels

​
DMs

DMs are the simplest way to work with Neo:
  • no mention is required
  • the conversation is naturally scoped
  • ideal for focused recon, research, and exploratory work

​
Channels

In channels, Neo is best used when the task belongs in a team conversation:
  • in public and private channels Neo only responds when explicitly mentioned with @Neo — plain channel messages are ignored even if the bot is in the channel
  • in group DMs (multi-person DMs) Neo also requires an @Neo mention to act
  • keep follow-up in the same thread
  • share findings where the rest of the team can see them
If Neo is disabled for a channel, users are guided toward admin settings or a DM instead.

​
Starting A Task

Slack works well for instructions such as:
  • “Run a full recon on example.com and look for exposed admin endpoints.”
  • “Review this code path for auth bypass or IDOR risk.”
  • “Research this CVE and tell me whether it is relevant to our stack.”
  • “Summarize the likely impact of this finding and what to check next.”
Neo can also work from attached files when the request needs more context.

​
Attachments And Files

When you attach files to your Slack request, Neo can use them as part of the task context. Good examples include:
  • logs
  • text summaries
  • code snippets
  • reports
  • exported findings or evidence files
For larger or more detailed outputs, Neo can return a concise Slack summary while attaching result files back into the thread.

​
What Results Look Like

Neo is designed to keep Slack readable:
  • it posts concise summaries in-thread
  • it highlights the key outcome instead of dumping raw output
  • it attaches files when the detailed output belongs outside the message body
  • it links the work back to Neo for deeper follow-up
Neo posting a Slack channel summary of a weekly security audit with attached JSON and Markdown reports

​
Task Visibility Is Team-Specific

Work started from Slack is not treated as a private task owned only by the person who sent the message.
  • each Slack workspace is connected to a specific Neo team
  • tasks created from Slack are stored in that Neo team context
  • visibility inside Neo follows the connected team’s access model
  • DMs keep the Slack conversation focused, but the resulting Neo task is still team-scoped

​
Continue In Neo

Slack is the fast front door, not the only surface. If the task becomes more complex, Neo continues to be available in the full Neo UI for:
  • full execution context
  • complete artifacts
  • longer follow-up tasks
  • reviewable task history

​
Stop A Running Task With Reactions

If Neo is still working on a request, you can abort the running task directly from the Slack thread by reacting with any of these:
Reaction-based abort is useful when:
  • the task is no longer needed
  • the target or scope changed
  • you realize the prompt needs to be corrected
  • you want to stop long-running work before it finishes
What to expect:
  • only a currently running Slack task is affected
  • Neo confirms the abort request in the thread
  • unrelated reactions do not stop work
If your workspace installed Neo before reaction-based abort shipped, reinstall or reconnect the app after the permissions update so the latest event subscription and scopes are active.

​
Example Prompts

Try prompts like these:
  • “Run a full recon on acme.com and summarize exposed services and likely risk areas.”
  • “Review this API design for authentication and authorization gaps.”
  • “Research CVE-2026-XXXX and tell me whether it matters for our environment.”
  • “Look through this attached report and summarize the issues worth validating first.”
  • “Continue the investigation from above and tell me what you would test next.”

​
Next