Hello, from what I understand Cody has recently received some kind of context caching/re-use update on VS Code? Currently, in a chat session if I add the current file to the context again, even if its contents have changed and don’t match the ones the LLM model is aware of, it appears that the old cached version is still used by the model, in turn leading to code generated based on old assumptions. Is anyone having a similar experience?
Hey @ernests Do you regulary push commits to the local main branch? If this does not resolve the issue, please open the command palette and search for “Cody: Update Search Index for …” to manually trigger the indexing for the current workspace.
This should resolve it finally.