"Contribute to Feast itself: setup, scoped tests, lint/type checks,
Scanned 9/8/2026
Install to Claude Code
npx -y skills add VectorSpaceLab/AREX-Skill --skill repo-development --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Repo Development?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/vectorspacelab-repo-development-ecb16325)More formats (shields.io, HTML) on the badges page.
---
name: repo-development
description: "Contribute to Feast itself: setup, scoped tests, lint/type checks,
docs/blog placement, protobufs, Go/Java/operator awareness, PR conventions,
and safe native verification selection."
disable-model-invocation: true
metadata:
disco-role: operating
license: Apache 2.0
---
# Feast Repo Development
Use this sub-skill when the user asks to modify Feast source code, add or reuse tests, run contributor checks, update documentation, regenerate protobufs, touch Go/Java/operator code, prepare a PR, or choose safe verification commands for a Feast repository change.
## Route First
- Use this sub-skill for Feast contributor setup, source-tree orientation, Makefile targets, `uv`/`pixi` workflows, scoped `pytest`, `ruff`, `mypy`, docs placement, protobuf generation, Go/Java commands, PR conventions, and safe native verification selection.
- Route end-user feature repository creation, `feast init`, `feature_store.yaml`, `feast apply`, and `feast plan` to `../feature-repos-and-cli/SKILL.md`.
- Route Feast object modeling with `Entity`, `Field`, `FeatureView`, transformations, or feature services to `../feature-definitions/SKILL.md`.
- Route retrieval, materialization, `push`, and online/historical feature API behavior to `../retrieval-and-materialization/SKILL.md`.
- Route feature servers, registry/offline servers, remote stores, TLS/auth/RBAC serving to `../servers-and-remote/SKILL.md`.
- Route optional backend/store/provider implementation design to `../integrations-and-extensibility/SKILL.md`; return here for Feast-core tests, docs, lint, and PR workflow.
## Fast Path
1. Confirm context: changed files, intended behavior, whether service-backed integration tests are safe, and whether protobufs, docs navigation, Go, Java, UI, or operator code are in scope.
2. Prefer targeted checks first: `uv run ruff check <python paths>`, `uv run ruff format <python paths>`, `uv run bash -c "cd sdk/python && mypy feast/<module>.py"`, and `uv run python -m pytest <specific tests> -v`.
3. Use `python scripts/select_feast_tests.py <changed paths...>` from this sub-skill to suggest small scoped checks without running them.
4. Escalate to Makefile targets only when needed: `make test-python-smoke`, `make test-python-unit`, `make lint-python`, `make mypy-full`, `make compile-protos-python`, `make protos`, `make build-go && make test-go`, or `make test-java`.
5. Avoid broad or service-backed integration suites unless explicitly authorized and prerequisites are available; record skipped unsafe native checks clearly.
6. For docs, update navigation and use the blog path rule: blog posts belong under `infra/website/docs/blog/`; other docs belong under `docs/` and should be discoverable from navigation.
## References
- `references/development-workflows.md` for setup, source-tree orientation, Makefile targets, PR conventions, and language/component awareness.
- `references/testing-guide.md` for safe test selection, unit vs integration boundaries, markers, and difficult change examples.
- `references/docs-and-protos.md` for docs placement, blog rules, protobuf generation, Go/Java/operator notes, and generated artifact checks.
- `references/troubleshooting.md` for install/import, optional extras, validation, CLI/API misuse, credentials, service prerequisites, and workflow failures.
- `scripts/select_feast_tests.py` for advisory scoped command suggestions from changed paths; it does not execute tests or require a Feast checkout.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!