{"conversation_id":"01ddc832-d30c-47fe-9459-8fd8169e6072"}
v0.2025.10.22.08.13.stable_01
gpt-5-high (create/edit tools)
#bug feedback
oblique bloom
mellow totemBOT
Hi @oblique bloom. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:
Based on the information in the knowledge sources, there are several known issues with GPT-5 (including high-reasoning variants) in Warp when using the create/edit tools:
-
Users have reported that attempts to edit or create files with GPT-5 often result in errors, such as gray error messages indicating that the code could not be applied. The model may report that it has made changes, but in reality, no edits are performed. This is often due to the tool requiring exact string matches, and if the file content differs even slightly (e.g., in whitespace or formatting), the operation fails. The model may not always recognize or report these failures accurately, leading to confusion and wasted tokens. This issue has been observed since version v0.2025.08.06.08.12.stable_02 and persists in later versions as well gray constant errors and token-price in GPT 5.
-
There are also reports that, after such errors, subsequent prompts may not work with any model until the session is reset, indicating a possible state corruption or unrecoverable error in the agent workflow Response stream finished unexpectedly and unable to recover.
-
The Warp team is aware of these issues and is actively investigating and working on fixes. In the meantime, users are encouraged to provide debugging IDs and logs to help with troubleshooting. You can gather the AI debugging ID by right-clicking on the AI conversation block and selecting "Copy debugging ID" Gathering AI debugging ID.