Monitor socials and flag comments, DMs and mentions worth answering
Agents are given access to social accounts to watch for comments, DMs and mentions and notify the user when one is worth a reply.
What they did
The user connects their social accounts to FanBase MCP, which gives an agent 20+ tools, and then uses it from ChatGPT, Claude Code, iMessage or another MCP client. The agent queries activity, looks at fan activity, drafts replies, and can send routine ones. It can run a scheduled morning inbox check and briefing. The user sets permissions, such as requiring approval before sending replies or publishing posts.
How it went
The page shows example runs, not measured results. In one, 3 priority conversations were left for approval and 8 or 11 routine ones were handled automatically. It gives no accuracy or time-saved figures.
Worth knowing
Set permissions deliberately. In the page's example, reading and drafting are allowed outright, while sending replies and publishing need approval, so the agent only flags and drafts until you loosen that.
Labels
Agents are given access to social accounts to watch for comments, DMs and mentions and notify the user when one is worth a reply.
The user sets permissions, such as requiring approval before sending replies or publishing posts.
The agent queries activity, looks at fan activity, drafts replies, and can send routine ones.
The user connects their social accounts to FanBase MCP, which gives an agent 20+ tools, and then uses it from ChatGPT, Claude Code, iMessage or another MCP client.
The “Material” label failed its quality check and is not used for selection.
Details
- Agents
- custom-built or unnamed agent
- Source
- other
- Source date
- Oct 11, 2026 · Submitted by @vibegoat
- Accounts
- 1
- In registry since
- Oct 12, 2026