Repository navigation
Conversation
Expose nanoattention/flint through the existing OpenAI-compatible local-provider path, with a validated loopback endpoint shared by chat and model discovery. This replaces manual router wiring that conflicts with reserved built-in IDs while keeping inference in the separately installed runtime. The provider fails closed without local credentials; no weights are downloaded or engine code vendored. Co-Authored-By: Codegraff <blackfloofie@codegraff.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What changed
nanoattention/flintas a built-in OpenAI-compatible local provider, with chat and model discovery sharing one validated loopback endpoint.GRAFF_NANOATTENTION_URLendpoint override and explicitGRAFF_CONTEXT; adds a conservative 32K default context row.Why
Previously, using this local runtime required a project router entry. That becomes invalid once
nanoattentionis built in because router configuration correctly rejects collisions with reserved provider IDs. Reusing Codegraff's existing OpenAI-compatible local-provider path adds native routing without inventing a new protocol.The inference runtime and model checkpoint remain separately installed prerequisites. This PR does not vendor engine code, bundle or download weights, modify global settings, or fall back to a cloud provider. Endpoint overrides are restricted to loopback so prompts cannot be redirected to a remote host through this setting.