What to check before sending an AI chat handoff
Review the destination, source scope, files, and next action before moving a conversation to another AI assistant.
Find and fix a missing piece 2 min read
Check that the source was captured and synced, then confirm the destination can read Dray through its own connection. Ask it to report an access failure instead of continuing from a guessed summary.
A destination can open with a well-formed request even when the source has not finished syncing. Check the source state first. Repeatedly sending the same prompt will not repair missing data or an authorization problem.
Next, confirm the destination account and tools. A connection in one assistant does not authorize another. If the client exposes a tool error, keep the sanitized error and the time of the request for diagnosis.
You can write a checked brief or paste an approved excerpt into the destination. Make clear that this is the supplied context, not a complete copy of the original chat. Leave out unrelated messages and files.
Do not use a public share link merely to get around a private connection failure. That changes the access boundary. Review what the link exposes and who can open it before choosing that route.
TAKE THIS WITH YOU
An illustrative template. Replace the details with material you are allowed to share.
Try to read the referenced source once. If access fails, report the failure and stop. Do not infer the contents from the conversation title. If the source is available, confirm its topic and the current decision before continuing with the task.
USING THIS WITH DRAY
Dray’s handoff requires a captured, synced source and a connected destination. It can include stored files in the source scope; review those before retrying.
Dray is in beta. Available sources and connections depend on your setup. Read the data notes before granting access.
Product details are grounded in the guides below. Example projects and prompts are illustrative, not customer results.