bazel-build-optimization

Define external dependencies in WORKSPACE.bazel using http_archive for aspect_rules_js and rules_python, then register toolchains and npm packages via npm_translate_lock. Create per-directory BUILD.bazel files with ts_project and py_library targets for fine-grained caching.

Add --remote_cache=grpcs://your-cache-endpoint to .bazelrc for caching, and --remote_executor with platform definitions for remote execution. Use named configs like build:remote-cache and build:ci so teams can opt in per environment.

Use bazel query "rdeps(//..., //libs/utils:utils)" to find reverse dependencies of a target. You can also run bazel query "deps(//target)" --output=graph piped to Graphviz dot to visualize the dependency graph.

Slow builds usually come from coarse-grained targets, missing cache hits, or oversized actions. Run bazel build with --profile=profile.json and analyze it with bazel analyze-profile, or use --execution_log_json_file to identify slow actions.

Yes, define a rule in a .bzl file using the rule() function with an implementation that calls ctx.actions.run to invoke a builder tool. Declare inputs, outputs, and attributes like dockerfile and base_image, then load the rule in BUILD files.

Avoid glob for deps since explicit dependency lists improve cache correctness and prevent accidental coupling. Glob is acceptable for srcs within a package, but cross-package dependencies should always be declared explicitly.