Pre-submit Checks
Describe the bug
Every time I start a new AI conversation, the agent completely ignores the established inbound rules provided by Warp.
If I manually remind the agent about the rules, it apologizes (e.g., "Oops, I missed a few of them") and begins following them. This indicates the agent has access to the rules, but isn't applying them at the initialization of a new chat.
While I am aware of the warp.dm workaround, I prefer to use the standard inbound rules framework as intended by the company. Having to manually remind the AI of its instructions at the start of every single session is highly disruptive to my workflow. Could you look into why these rules aren't being automatically enforced at the start of new threads?
To reproduce
Whenever I start a new agent to edit a large codebase, it forgets the inbound instructions.
Expected behavior
The agent needs to process the inbound rules first before attempting to edit the code.
Screenshots, videos, and logs
Operating system (OS)
Linux
Operating system and version
fedora 44
Shell Version
fish
Current Warp version
No response
Regression
No, this bug or issue has existed throughout my experience using Warp
Recent working Warp date
No response
Additional context
No response
Does this block you from using Warp daily?
Yes, this issue prevents me from using Warp daily.
Is this an issue only in Warp?
Yes, I confirmed that this only happens in Warp, not other terminals.
Warp Internal (ignore): linear-label:b9d78064-c89e-4973-b153-5178a31ee54e
None
Pre-submit Checks
Describe the bug
Every time I start a new AI conversation, the agent completely ignores the established inbound rules provided by Warp.
If I manually remind the agent about the rules, it apologizes (e.g., "Oops, I missed a few of them") and begins following them. This indicates the agent has access to the rules, but isn't applying them at the initialization of a new chat.
While I am aware of the warp.dm workaround, I prefer to use the standard inbound rules framework as intended by the company. Having to manually remind the AI of its instructions at the start of every single session is highly disruptive to my workflow. Could you look into why these rules aren't being automatically enforced at the start of new threads?
To reproduce
Whenever I start a new agent to edit a large codebase, it forgets the inbound instructions.
Expected behavior
The agent needs to process the inbound rules first before attempting to edit the code.
Screenshots, videos, and logs
Operating system (OS)
Linux
Operating system and version
fedora 44
Shell Version
fish
Current Warp version
No response
Regression
No, this bug or issue has existed throughout my experience using Warp
Recent working Warp date
No response
Additional context
No response
Does this block you from using Warp daily?
Yes, this issue prevents me from using Warp daily.
Is this an issue only in Warp?
Yes, I confirmed that this only happens in Warp, not other terminals.
Warp Internal (ignore): linear-label:b9d78064-c89e-4973-b153-5178a31ee54e
None