mirror of
https://github.com/meta-llama/llama-stack.git
synced 2025-12-11 19:56:03 +00:00
0085db2c09
Backport UV index configuration fixes from release-0.3.x that enable resolution of RC dependencies from test.pypi while keeping PyPI as the primary index. Key changes: - Fix install-llama-stack-client action to use test.pypi as EXTRA index (not primary) to prevent UV from looking for all packages there first - Add uv-run-with-index.sh wrapper that auto-detects release branches and sets UV env vars for local pre-commit hooks - Update pre-commit hooks to use the wrapper for all uv commands - Add UV_INDEX_STRATEGY=unsafe-best-match to allow checking multiple indexes - Pass UV env vars as Docker build args and use them inline only for llama-stack installation to avoid affecting distribution deps - Export UV env vars to GITHUB_ENV in setup-runner for cross-step persistence This infrastructure allows future release branches to work correctly without manual UV configuration. |
||
|---|---|---|
| .. | ||
| github | ||
| telemetry | ||
| check-init-py.sh | ||
| check-workflows-use-hashes.sh | ||
| diagnose_recordings.py | ||
| distro_codegen.py | ||
| docker.sh | ||
| gen-changelog.py | ||
| gen-ci-docs.py | ||
| generate_prompt_format.py | ||
| get_setup_env.py | ||
| install.sh | ||
| integration-tests.sh | ||
| normalize_recordings.py | ||
| provider_codegen.py | ||
| run-ui-linter.sh | ||
| unit-tests.sh | ||
| uv-run-with-index.sh | ||