server · tiles · Tangled

A local-first, collaborative AI assistant that works for you. Built with open models and decentralized protocols. www.tiles.run

tiles server

3 folders · 13 files

· Fix/mtp opt in (#195) · 2mo ago

Fix/mtp opt in (#195)

Fix/mtp opt in (#195)

Fix/mtp opt in (#195)

feat: Python server for running mlx models

fix: Added exclude-newer to venvstack to prevent accidental lib upgrade that broke the build

Fix/mtp opt in (#195)

Some fixes related to tool call id, added a cuda13 script for Arch Linux based distros, and more test cases

More flags in llama server for mac os compatibility

fix: tool calling 422s on Linux builds due to openresponses-types version split (#188)

pyrightconfig.json

feat: Better of support of responses api with built harmony support

reasoning_utils.py

requirements-linux.txt

llama-server implementation

requirements-macos.txt

llama-server implementation

feat: first stab at responses api

fix: tool calling 422s on Linux builds due to openresponses-types version split (#188)

fix: tool calling 422s on Linux builds due to openresponses-types version split (#188)