Letting an AI agent post to Threads without letting it go rogue

Description and use

An AI agent can post to Threads with mandatory human approval, content guardrails, and rate‑limit compliance.

Problem

Ensuring an AI agent can schedule and publish Threads posts while preventing it from posting unapproved content.

Outcome

Agents can now post to Threads with explicit human approval, content guardrails, and rate‑limit compliance, reducing the risk of rogue posts.

Workflow

• 1. Sign up at mindthread.tw and connect a Threads account via OAuth.
• 2. Create an API key (test or live).
• 3. Use the CLI to schedule a post with mindthread post "text" --account @handle --at datetime.
• 4. For agents, use --json and monitor exit codes (0‑5).
• 5. Point an MCP‑compatible client to the MindThread MCP endpoint with the key.
• 6. Configure key permissions (read, write drafts, schedule, publish, etc.) to allow, ask, or deny.
• 7. When a key is set to ask, the agent receives an approval_required link and waits for user confirmation on mindthread.tw.
• 8. Use mindthread approval <id> or --remind to check status.
• 9. Manage pending reply drafts with mindthread patrol pending and mindthread patrol approve.
• 10. Handle rate limits via exit code 5 or 429 responses.
• 11. Use mindthread events --follow for event feed or webhooks for real‑time updates.

Evidence

Letting an agent post is easy. Making sure it can’t post something you didn’t approve took most of the work.

Source

Each permission on a key (read, write drafts, schedule, publish now, send replies, turn cloud patrol on, connect accounts, loosen guardrails, webhooks) is set to allow, ask or deny.

Source

With ask, the request does not run. The agent gets approval_required with a link to hand you, and the action runs only after you confirm on mindthread.tw.

Source

In October 2026 the free plan is 1 account, 3 posts a day and 50 calls a day per live key, test keys share a 500-a-day sandbox, and paid plans raise the daily pool; a per-minute limit applies too.

Source