Giving an AI agent memory of our Twist threads: building a Twist connector for cognee
说明与用途
为 cognee 构建了 Twist 连接器,将长时间的 Twist 讨论同步为单一 AI 内存文档,实现无需打开 Twist 即可进行自然语言查询和准确回忆决策。
问题
Twist threads are long and slow, making it hard to retrieve decisions later; search finds the thread but not the answer.
结果
The connector successfully ingests Twist threads as single documents, handles deletions, and allows accurate recall of decisions without opening Twist.
工作流
- 1. Set up a Twist workspace with a #launch channel and a thread titled "Checkout redesign". 2. Sync the Twist workspace to cognee using the cognee‑community‑connector‑twist. 3. Ask a natural‑language question via cognee’s recall API. 4. Receive the best matching passage from the synced data.
- 5. Delete a thread in Twist, sync again, and confirm the deleted thread no longer appears in recall results.
证据
Twist is built for long, slow conversations. Someone opens a thread with a proposal, people reply over two or three days, and somewhere around the fifth comment the actual decision gets made.
You point it at a Twist workspace, it pulls threads and channels into cognee, and then you can ask questions in plain language.
After a sync, this is the document cognee receives for that thread. This is real output from my workspace: ...
These are the results I got on my workspace. I ran this one without an LLM key, in cognee's keyless mode, so recall returns the best matching passage instead of a written answer:
One document per thread, not per comment.
It goes through dlt, with "merge".
So last_updated_ts misses most edits. There is a version field that catches nearly everything...
- 记录日期: 2026-11-01T23:44:02.510Z