Skip to content

Add local Liquid coding agent provider - #1

Draft
hudsonaikins wants to merge 13 commits into
mainfrom
feat/local-liquid-coding-agent
Draft

Add local Liquid coding agent provider#1
hudsonaikins wants to merge 13 commits into
mainfrom
feat/local-liquid-coding-agent

Conversation

@hudsonaikins

@hudsonaikins hudsonaikins commented Aug 30, 2026

Copy link
Copy Markdown
Owner

Summary

Adds local Liquid AI coding-agent support to this fx fork.

  • Routes LM Studio through OpenAI-compatible local Chat Completions and Responses handling.
  • Keeps Graphify MCP retrieval and local permission review in the existing FX runtime.
  • Includes Unsloth QLoRA and export scripts for private local experiments.
  • Documents 128K local context setup and per-invocation model switching.

Training trajectories, adapters, checkpoints, and compiled caches stay outside the repository tip. Training scripts read local paths through FX_TRAJECTORY_DATASET and FX_UNSLOTH_OUTPUT.

Validation

  • Zig 0.16.0 Linux cross-build passed.
  • Built ./zig-out/bin/fx --help exited 0 with empty stderr.
  • Full CI passed for the preceding code commit; checks for this documentation-only commit are pending.

@hudsonaikins hudsonaikins added the type: feature Adds a new user-facing capability label Aug 30, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

type: feature Adds a new user-facing capability

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants