git 5.0.2 → 5.0.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 94d97e48dd1d4b34affe837d38258f7b0f7edf02338d0e0def333ac80e147244
-  data.tar.gz: 5367c11505c46d9745d492a97671346531fddfdc3649b445bae9eb2636f87202
+  metadata.gz: 6dcf907e2c0ccb6d916a9aa1d43f6c768e49555465a484e0231d919cb889b889
+  data.tar.gz: 5cce608675a41031f79318065e8c11125f9f23a154ab2ae6f578192010b04367
 SHA512:
-  metadata.gz: 4e404e6b31961cf3c71e391ce4898436f07de8dd2ffc37947281f103513241c531e21aa799ae32296ec6ef47d2b1dad465205fd3b0936ee104b6774870805391
-  data.tar.gz: 778bbd70beb547b4574e8d361acc2873320b42bd4b5f190012309a3b90da72d55468a8c5b4adf87a3ed773ab08945e7ac81b3d282c86bb8bda64821282a3c101
+  metadata.gz: 42dbbce6613ecd406e64771bb03d85416a4eae0fe12183fee03640292f5f18700175788c35c55836ad2d572bde6857421ed6cae8f9d7520dcd8bd12b498ae809
+  data.tar.gz: d55dd4bd23e700aacd23d8b850adada2e0e3a8f90176515e5de3b5213e43282ac5e8e105d500243bc54a0655472368e00becf4b6c01e59e4baf842223f8f89a7

data/.claude/commands/address-copilot-reviews.md

@@ -0,0 +1,14 @@
+---
+description: Address all unresolved Copilot review threads on the active pull request
+argument-hint: "[PR number — defaults to the PR for the current branch]"
+---
+
+Follow the shared agent prompt in
+`.github/prompts/iteratively-address-copilot-reviews.prompt.md` exactly. That file
+is the canonical instructions for this task and is shared with other AI agents; read
+it before doing anything else.
+
+@.github/prompts/iteratively-address-copilot-reviews.prompt.md
+
+If `$ARGUMENTS` is non-empty, treat it as the pull request number to operate on
+instead of detecting the PR from the current branch.

data/.claude/settings.json

@@ -0,0 +1,16 @@
+{
+  "$schema": "https://json.schemastore.org/claude-code-settings.json",
+  "hooks": {
+    "SessionStart": [
+      {
+        "hooks": [
+          {
+            "type": "command",
+            "command": "\"$CLAUDE_PROJECT_DIR/.github/hooks/run-bin-setup-once.sh\"",
+            "timeout": 900
+          }
+        ]
+      }
+    ]
+  }
+}

data/.claude/skills

@@ -0,0 +1 @@
+../.github/skills

data/.github/copilot-instructions.md

@@ -22,18 +22,6 @@ fine to use `#` in footer values such as `Closes: #999` or `Refs: #999`.
 - **RuboCop:** Use `RuboCop` for prose and `rubocop` for CLI or gem names. Never use
   `Rubocop`.
 
-## Project Commands
-
-| Purpose | Command |
-| --- | --- |
-| First-time project or new worktree setup | `bin/setup` |
-| Run all tests and linters (CI-equivalent) | `bundle exec rake` |
-| Run all RSpec tests | `bundle exec rake spec` |
-| Run RSpec unit tests | `bundle exec rake spec:unit` |
-| Run RSpec integration tests | `bundle exec rake spec:integration` |
-| Run a specific RSpec spec | `bundle exec rspec <path>` |
-| Run linters | `bundle exec rake rubocop yard build` |
-
 ## Branch & PR Strategy
 
 | Target | When |

data/.github/hooks/run-bin-setup-once.sh

@@ -1,12 +1,18 @@
 #!/usr/bin/env bash
 set -euo pipefail
 
+# This hook is shared by every agent configured for this repo (GitHub Copilot via
+# .github/hooks/, Claude Code via .claude/settings.json), so do not assume the
+# caller's working directory.
+command -v git >/dev/null 2>&1 || exit 0
+cd "$(dirname "${BASH_SOURCE[0]}")/../.." || exit 0
+
 # Skip if this repo does not use bin/setup.
 [[ -x "./bin/setup" ]] || exit 0
 
-# Keep setup idempotent per-worktree.
-command -v git >/dev/null 2>&1 || exit 0
-marker="$(git rev-parse --git-path copilot-bin-setup.done 2>/dev/null)" || exit 0
+# Keep setup idempotent per-worktree. The marker is shared across agents so that
+# cloning and opening the repo in two different tools does not run setup twice.
+marker="$(git rev-parse --git-path agent-bin-setup.done 2>/dev/null)" || exit 0
 [[ -f "$marker" ]] && exit 0
 
 ./bin/setup

data/.github/pull_request_template.md

@@ -16,4 +16,6 @@ Review our [guidelines for contributing](https://github.com/ruby-git/ruby-git/bl
 - [ ] I ran `bundle exec rake` locally on this branch and it passed (see
   [Local development setup](../CONTRIBUTING.md#local-development-setup)).
 - [ ] Tests added/updated as needed.
+- [ ] `bundle exec rake spec:unit` reports 100% line and branch coverage (see
+  [Test coverage policy](../CONTRIBUTING.md#test-coverage-policy)).
 - [ ] Documentation updated where applicable (README and/or YARD).

data/.github/skills/command-implementation/REFERENCE.md

@@ -908,7 +908,7 @@ For migration PRs, verify process constraints:
 - each slice is independently revertible
 - refactor-only changes are not mixed with unrelated behavior changes
 - quality gates pass for the slice — discover tasks via
-  `bundle exec ruby -e "require 'rake'; load 'Rakefile'; puts Rake::Task['default:parallel'].prerequisites"`
+  `bundle exec ruby -e "require 'rake'; load 'Rakefile'; puts Rake::Task['default'].prerequisites"`
   and run each individually via `bundle exec rake <task>`, fixing failures before
   advancing
 

data/.github/skills/command-implementation/SKILL.md

@@ -188,11 +188,11 @@ This skill supports three modes. Determine which mode applies before starting:
      requirements](REFERENCE.md#phased-rollout-requirements).
 
 5. **Run quality gates** — discover the prerequisite tasks for
-   `default:parallel` and run them sequentially, fixing failures before
+   `default` and run them sequentially, fixing failures before
    advancing:
 
    ```bash
-   bundle exec ruby -e "require 'rake'; load 'Rakefile'; puts Rake::Task['default:parallel'].prerequisites"
+   bundle exec ruby -e "require 'rake'; load 'Rakefile'; puts Rake::Task['default'].prerequisites"
    ```
 
    Run each listed task **individually** in order via `bundle exec rake <task>`

data/.github/skills/facade-implementation/SKILL.md

@@ -224,11 +224,11 @@ This skill supports three modes. Determine which mode applies before starting:
    - has tests that verify call-shape compatibility when classification is
      `legacy-contract` (positional hash and/or keyword-arg / `**opts` forms where required)
 
-5. **Run quality gates** — discover the prerequisite tasks for `default:parallel`
+5. **Run quality gates** — discover the prerequisite tasks for `default`
    and run them sequentially, fixing failures before advancing:
 
    ```bash
-   bundle exec ruby -e "require 'rake'; load 'Rakefile'; puts Rake::Task['default:parallel'].prerequisites"
+   bundle exec ruby -e "require 'rake'; load 'Rakefile'; puts Rake::Task['default'].prerequisites"
    ```
 
    Run each listed task **individually** in order via `bundle exec rake <task>`

data/.github/skills/rspec-unit-testing-standards/SKILL.md

@@ -68,9 +68,13 @@ Adoption and enforcement notes:
 - Apply these rules as hard requirements for new and modified unit specs.
 - Legacy specs may violate some rules; treat those as incremental cleanup work.
 - Branch and line coverage are both reported by SimpleCov in this repository.
-- `coverage_threshold: 100` is configured, but `fail_on_low_coverage` is currently
-  `false`; enforce Rule 21 during review by checking the coverage report until
-  strict failure is enabled.
+- `minimum_coverage: { line: 100, branch: 100 }` is configured and enforced: a full
+  `rake spec:unit` run fails when either threshold is missed, and CI fails the pull
+  request with it. Rule 21 is a build gate, not just a review check. See the
+  [Test coverage policy](../../../CONTRIBUTING.md#test-coverage-policy) in
+  `CONTRIBUTING.md` for the full policy.
+- Focused runs (`SPEC=<glob> rake spec:unit`, or `rspec <file>`) report coverage but
+  do not fail on it, since they load all of `lib/` while exercising only a slice.
 
 ## Related skills
 
@@ -484,10 +488,50 @@ Every conditional path through the public interface must be exercised by at leas
 one example. If a branch cannot be reached through the public interface, that is a
 design smell.
 
+This is enforced by the build: a full `rake spec:unit` run fails below 100% line or
+branch coverage, and the failure output names each uncovered line and branch.
+
+When a branch is hard to cover, apply these in order:
+
+1. **Reach it through the public interface** — if it is reachable, test it.
+2. **Delete it** — a branch unreachable through the public interface is usually dead
+   code. Removing it is preferable to excluding it.
+3. **Exclude it with `# simplecov:disable`** — the last resort, per the exception below.
+
 > **Exception:** Defensive guards that require breaking OS-level invariants to reach
 > (e.g., `raise "unreachable"` that would require triggering out-of-memory) may be
-> excluded. Mark them explicitly with `# :nocov:` and a brief comment explaining why
-> — never leave branches silently uncovered.
+> excluded. Mark them explicitly with a `# simplecov:disable` directive carrying a
+> reason — never leave branches silently uncovered. Cover the smallest possible span,
+> and expect a reviewer to challenge it. `lib/` currently contains no coverage
+> directives.
+
+Prefer the inline form for a single line. It applies only to its own line and needs no
+matching `enable`, so a region can never be left open by accident:
+
+```ruby
+raise 'unreachable' # simplecov:disable defensive guard; only reachable on OOM
+```
+
+Use the block form only for a genuine span, and close it explicitly:
+
+```ruby
+# simplecov:disable branch platform-specific fallback; not reachable on MRI
+...
+# simplecov:enable branch
+```
+
+Always name the narrowest criterion that solves the problem — `line`, `branch`,
+`method`, or a comma-separated combination — and write the reason after it, so the
+required justification lives in the directive.
+
+> **Review check:** verify the criterion is spelled exactly. A word SimpleCov does not
+> recognize is parsed as free-form reason text, which silently widens the directive to
+> all three criteria rather than failing.
+
+> **Coverage is a floor on evidence, not a proof of correctness.** Never write an
+> example whose only purpose is to execute a line. Such a test turns the report green
+> while violating [Rule 24](#rule-24-must-assert-observable-behavior-not-implementation-details),
+> and must be rejected in review.
 
 ### Rule 22 (MUST): Error assertions must specify both the error class and a message pattern
 
@@ -614,7 +658,9 @@ After writing or modifying tests, verify compliance before finishing:
 
 1. **Run the specs:** `bundle exec rspec spec/unit/path/to_spec.rb`
 2. **Check branch coverage** meets 100% (Rule 21) — open `coverage/index.html` and
-   confirm no uncovered branches in the class under test.
+   confirm no uncovered lines or branches in the class under test. A focused run
+   reports coverage but does not fail on it; `bundle exec rake spec:unit` is the
+   authority and will fail below 100%, naming each uncovered line and branch.
 3. **Re-check MUST rules.** Scan the spec against every MUST rule. Fix violations.
 4. **Run in random order** (Rule 27): `bundle exec rspec spec/unit/path/to_spec.rb --order rand`
 

data/.github/workflows/continuous_integration.yml

@@ -53,5 +53,17 @@ jobs:
         run: bundle exec rake default
         timeout-minutes: 15
 
+      # The unit suite enforces 100% line and branch coverage (see the "Test coverage
+      # policy" section of CONTRIBUTING.md). The failure log names the uncovered lines
+      # and branches; this artifact adds the browsable HTML report for the same run.
+      - name: Upload Coverage Report
+        if: failure()
+        uses: actions/upload-artifact@v7
+        with:
+          name: coverage-ruby-${{ matrix.ruby }}-${{ matrix.operating-system }}
+          path: coverage/
+          if-no-files-found: ignore
+          retention-days: 7
+
       - name: Test Gem
         run: bundle exec rake test:gem

data/.gitignore

@@ -25,3 +25,6 @@ git-reference/
 git-versions/
 
 tmp
+
+# per-developer Claude Code overrides (shared config is in .claude/settings.json)
+.claude/settings.local.json

data/.release-please-manifest.json

@@ -1,3 +1,3 @@
 {
-  ".": "5.0.2"
+  ".": "5.0.3"
 }

data/.rspec

@@ -1,3 +1,2 @@
 --require spec_helper
---format Fuubar
 --color

data/CHANGELOG.md

@@ -5,6 +5,47 @@
 
 # Change Log
 
+## [5.0.3](https://github.com/ruby-git/ruby-git/compare/v5.0.2...v5.0.3) (2026-08-07)
+
+
+### Bug Fixes
+
+* Drop unsupported -- separator from remote remove/rename commands ([8ec466f](https://github.com/ruby-git/ruby-git/commit/8ec466f8b282a5e94abd7d499bc26a92d94b2fc3))
+* Remove unreachable case branch in Patch#apply_file_mode ([896fc49](https://github.com/ruby-git/ruby-git/commit/896fc49e3c2c03c401fe1b213c3f75bb40730521))
+* Remove unreachable nil check in Git::Parsers::Grep.parse ([74e919a](https://github.com/ruby-git/ruby-git/commit/74e919a401126ad4415615f92cc9597e966239c6))
+
+
+### Other Changes
+
+* Add Claude Code support alongside GitHub Copilot ([37f7d6a](https://github.com/ruby-git/ruby-git/commit/37f7d6a14c6d53bcf5b9644b8023d323290a5977))
+* Add unit spec for Git::EncodingUtils ([5f0f10e](https://github.com/ruby-git/ruby-git/commit/5f0f10e4db78324bbe453de49598564a25f3428a))
+* Consolidate parallel and serial spec rake tasks ([49616b1](https://github.com/ruby-git/ruby-git/commit/49616b12f52c6941a8617afbf7c56e7d7d5fe0f6))
+* Cover :index option in Git::Factories#worktree_open_options_after_init ([efeba22](https://github.com/ruby-git/ruby-git/commit/efeba2294bb05327489a6523c46f26dc7392a84e))
+* Cover #path when both src and dst are nil ([3a1a05a](https://github.com/ruby-git/ruby-git/commit/3a1a05a9be13057a1fc304df5008866634179124))
+* Cover edge-case branches in Git::Parsers::Diff ([38d8b3f](https://github.com/ruby-git/ruby-git/commit/38d8b3f88f4f76a65776e9f6b8b87c6bb839e710))
+* Cover Errno::ESRCH timeout-race rescue in run_process_executer ([899bf36](https://github.com/ruby-git/ruby-git/commit/899bf36399489038524cecf1d7ac2c606df03bde))
+* Cover Git::Branch#merge ([71c0ea3](https://github.com/ruby-git/ruby-git/commit/71c0ea39d4c786a38245a8fb95fd93235ee03903))
+* Cover Git::Branch#stashes ([27ba130](https://github.com/ruby-git/ruby-git/commit/27ba130a6cd3ef78171d1bb20230459c3f678a3a))
+* Cover Git::Diff#name_status ([fa2ee98](https://github.com/ruby-git/ruby-git/commit/fa2ee98fd4613177843d285f7bc4a2ffb44ca79c))
+* Cover Git::Factories#parse_clone_stderr unparseable-output branch ([65aa9b9](https://github.com/ruby-git/ruby-git/commit/65aa9b9318c7f8a8986c2aaae7dd0af0aeec4d69))
+* Cover Git::Parsers::CatFile.each_header without trailing separator ([b7a9ed9](https://github.com/ruby-git/ruby-git/commit/b7a9ed93e03dcbd22a4fb25e9bb7b9c49647262d))
+* Cover Git::Parsers::LsTree.parse with no tab-separated filename ([98b9ebf](https://github.com/ruby-git/ruby-git/commit/98b9ebf03c49f51c76b2e56dd670a9f68f220da8))
+* Cover Git::Parsers::Remote.apply_pair unrecognized-variable branch ([03b10f6](https://github.com/ruby-git/ruby-git/commit/03b10f618a2f012b578ab995120cedbcd96ed8d1))
+* Cover Git::Stashes#save ([7bca04c](https://github.com/ruby-git/ruby-git/commit/7bca04c53e64c747627784773d0c03ad0b9ff2e3))
+* Cover Git.const_missing fallback branch for unrelated constants ([05c3c43](https://github.com/ruby-git/ruby-git/commit/05c3c4350e18bf1403fa314902708cfc40848c70))
+* Cover Git.export ([4bd7f58](https://github.com/ruby-git/ruby-git/commit/4bd7f58432a19385ff55aecca9e89e2d9a96fa4a))
+* Cover normalize_fetch_keys with a String option key ([47fb7fe](https://github.com/ruby-git/ruby-git/commit/47fb7fe061df9177a26ec7ada90b35b902ffc1df))
+* Disable coverage for spec:integration, add SPEC filtering, fix formatters ([cb53d03](https://github.com/ruby-git/ruby-git/commit/cb53d03f8b671873f24368aadcafec638de32345))
+* Document the project test coverage policy ([d378c1e](https://github.com/ruby-git/ruby-git/commit/d378c1e8b75e0c4209e17cc040ed36eb93053797))
+* Enforce 100% unit test coverage on pull requests ([a446098](https://github.com/ruby-git/ruby-git/commit/a446098af548039b2c8b68bbb6562a53cb79fe40))
+* Guard checkout-index nonexistent-path spec for git < 2.30.0 ([cf212a9](https://github.com/ruby-git/ruby-git/commit/cf212a98b41218a7940ce54aa9b7c56a630412c3))
+* Guard worktree repair spec for git < 2.29.0 ([c527fb1](https://github.com/ruby-git/ruby-git/commit/c527fb15c2f21f9be5913192d43c847cf49681eb))
+* **merge:** Fix incorrect version-gated failure expectation in quit_spec ([24c5ba4](https://github.com/ruby-git/ruby-git/commit/24c5ba4ce77b02413485d226a94b283f6af9d110))
+* Read the gem version without loading lib/git/version.rb ([45c4b14](https://github.com/ruby-git/ruby-git/commit/45c4b14e944fcb43bda800823e377cf954ce2d3b))
+* Require git 2.39.0 for maintenance register/unregister --config-file specs ([5b1fd52](https://github.com/ruby-git/ruby-git/commit/5b1fd527c39e4b3f1796851795a44694d94f4402))
+* Skip get-urlmatch scope spec on git < 2.42.0 ([e1ede7a](https://github.com/ruby-git/ruby-git/commit/e1ede7adddea28b654a183c618fc2336b5ffb83b))
+* Upgrade simplecov and simplecov-rspec to ~> 1.0 ([65e92b9](https://github.com/ruby-git/ruby-git/commit/65e92b9a9a621435efbf2a0855d48c47bb62425d))
+
 ## [5.0.2](https://github.com/ruby-git/ruby-git/compare/v5.0.1...v5.0.2) (2026-08-02)
 
 

data/CLAUDE.md

@@ -0,0 +1,11 @@
+# CLAUDE.md
+
+This file exists so Claude Code picks up the same project instructions as every
+other AI coding agent used in this repository.
+
+**Do not add project guidance here.** The canonical instructions live in
+[`.github/copilot-instructions.md`](.github/copilot-instructions.md) and are
+imported below. Edit that file so all agents stay in sync. Reserve this file for
+guidance that applies *only* to Claude Code.
+
+@.github/copilot-instructions.md

data/CONTRIBUTING.md

@@ -20,6 +20,7 @@
   - [Before requesting review](#before-requesting-review)
 - [Branch strategy](#branch-strategy)
 - [AI-assisted contributions](#ai-assisted-contributions)
+  - [Agent configuration](#agent-configuration)
   - [Agent skills](#agent-skills)
 - [Design philosophy](#design-philosophy)
 - [Layered architecture](#layered-architecture)
@@ -39,6 +40,7 @@
     - [What to know about Conventional Commits](#what-to-know-about-conventional-commits)
     - [Issue and PR references](#issue-and-pr-references)
   - [Testing guidelines](#testing-guidelines)
+    - [Test coverage policy](#test-coverage-policy)
     - [Unit tests vs Integration tests](#unit-tests-vs-integration-tests)
 - [Building a specific version of the Git command-line](#building-a-specific-version-of-the-git-command-line)
   - [Install pre-requisites](#install-pre-requisites)
@@ -215,6 +217,27 @@ and "CI is green" are not substitutes for the submitter running
 [the local validation step](#contributor-validation-policy) themselves; CI is a
 backstop, not a primary validation surface.
 
+### Agent configuration
+
+Agent configuration is shared: each piece of guidance is stored once and surfaced to
+every supported agent.
+
+| Content | Canonical location | Also read by |
+| --- | --- | --- |
+| Project instructions | [`.github/copilot-instructions.md`](.github/copilot-instructions.md) | Claude Code, via an import in [`CLAUDE.md`](CLAUDE.md) |
+| Skills | [`.github/skills/`](.github/skills/) | Claude Code, via the `.claude/skills` symlink |
+| Prompts | [`.github/prompts/`](.github/prompts/) | Claude Code, via wrappers in `.claude/commands/` |
+| Setup hook | [`.github/hooks/run-bin-setup-once.sh`](.github/hooks/run-bin-setup-once.sh) | Claude Code, via `.claude/settings.json` |
+
+Always edit the canonical file. The Claude Code side is a pointer in every case, so
+changes reach both agents without a sync step.
+
+One caveat: `.claude/skills` is a committed symlink. Git for Windows only
+materializes symlinks when `core.symlinks` is enabled (which requires Developer Mode
+or an elevated shell). Without it, Windows contributors get a plain text file there
+and Claude Code silently loads no skills; either enable symlinks or point your agent
+at [`.github/skills/`](.github/skills/) directly. Copilot is unaffected.
+
 ### Agent skills
 
 If you use an AI coding agent that understands repository skills, the
@@ -904,12 +927,138 @@ process.stdin.on('end', () =>
   local working copy.
 - Test runs are covered by SimpleCov by default. Set `COVERAGE=false` (or `0`/`no`/
   `off`) to skip coverage, e.g. `COVERAGE=false bundle exec rake spec`.
+  `rake spec:integration` always disables coverage, regardless of `COVERAGE`:
+  integration tests aren't meant to be exhaustive, so tracking their coverage would
+  misleadingly suggest that low integration coverage is a problem to fix.
+- `rake spec:integration` runs in parallel (via `parallel_tests`) on MRI. Set
+  `PARALLEL_TESTS=false` (or `0`/`no`/`off`) to force serial execution, e.g.
+  `PARALLEL_TESTS=false bundle exec rake spec:integration`. A run narrowed by `SPEC`
+  to a single spec file always runs serially — there is nothing to divide across
+  workers, and serial execution gives per-example (documentation) output.
+- Set `SPEC=<glob>` to run specific files instead of a task's whole directory, e.g.
+  `SPEC=spec/unit/git/version_spec.rb bundle exec rake spec:unit`.
+- Each task runs only the matches that live under its own directory, so one glob
+  spanning `spec/unit/` and `spec/integration/` can drive `bundle exec rake spec` and
+  exercise both layers of an area in a single command:
+
+  ```bash
+  # Unit and integration specs for the add command
+  SPEC=spec/**/git/commands/add_spec.rb bundle exec rake spec
+
+  # Unit and integration specs for every command class
+  SPEC=spec/**/git/commands bundle exec rake spec
+  ```
+
+  The glob is expanded by Rake, not the shell, so `**` works the same in any shell.
+  A task whose directory contains none of the matches is skipped with a message —
+  `SPEC=spec/unit/...` on `rake spec` runs the unit specs and skips
+  `spec:integration`. A glob matching nothing anywhere fails the task outright.
 
 This project uses **RSpec** (`spec/`) as its sole test framework. Structure,
 naming, setup, stubbing, and coverage rules for unit specs are defined in the
 [`rspec-unit-testing-standards`](.github/skills/rspec-unit-testing-standards/SKILL.md)
 skill — follow it when writing or reviewing specs under `spec/unit/`.
 
+#### Test coverage policy
+
+**Every pull request to `main` must keep `bundle exec rake spec:unit` at 100% line
+coverage and 100% branch coverage of `lib/`.** CI fails the build when it drops
+below either threshold.
+
+This is enforceable without being onerous because unit coverage in this project is
+deterministic: `lib/` has no Ruby-version, Ruby-engine, or platform conditionals, and
+no unit spec is conditionally skipped. Every supported MRI runtime measures exactly
+the same lines and branches, so a coverage failure is always something the pull
+request introduced.
+
+What the policy does and does not cover:
+
+- **Scope is the unit suite on MRI.** Integration specs are deliberately not
+  exhaustive and are not measured (`rake spec:integration` always disables coverage).
+  JRuby and TruffleRuby do not produce reliable coverage data and are not measured.
+- **Enforcement applies to full-suite runs.** A focused run
+  (`SPEC=<glob> bundle exec rake spec:unit`, or `bundle exec rspec <file>`) still
+  reports coverage but will not fail on it, so the usual edit-test loop is unaffected.
+  Set `FAIL_ON_LOW_COVERAGE=true` to force enforcement on for a focused run.
+- **A focused run lists gaps only in the code it is about.** The reported percentage is
+  always for the whole of `lib/`, but the list of uncovered lines and branches is scoped
+  to the files the run tests — the classes it describes, plus the `lib/` file each spec
+  file mirrors. So a focused run answers "is what I just changed fully covered?" without
+  waiting for CI:
+
+  ```text
+  Reporting uncovered lines and branches for 1 of 225 files.
+
+  No uncovered lines and branches in this file.
+  ```
+
+  Ignore the percentage on a focused run; it is low because `spec_helper` loads all of
+  `lib/`, not because anything is wrong. Set `LIST_UNCOVERED_FILES=all` to list gaps for
+  every file instead.
+- **The thresholds do not move.** Do not lower `minimum_coverage` and do not add a
+  SimpleCov filter to exclude a file. The only sanctioned escape hatch is a
+  `# simplecov:disable` directive.
+- **Every file under `lib/` is measured.** SimpleCov only tracks files loaded after it
+  starts, so a file loaded earlier is silently absent from the report rather than
+  counted as uncovered. This is why `git.gemspec` reads the version string out of
+  `lib/git/version.rb` instead of `require`ing it: the `Gemfile` uses `gemspec`, so
+  requiring it there would load it on every `bundle exec`, before SimpleCov starts.
+  Keep new code out of the load path that runs ahead of `spec_helper`.
+
+When a branch is hard to cover, apply these in order:
+
+1. **Reach it through the public interface.** If a branch is reachable, test it.
+2. **Delete it.** A branch that cannot be reached through the public interface is
+   usually dead code, and removing it is preferable to excluding it. See commit
+   `74e919a4` ("fix: remove unreachable nil check in `Git::Parsers::Grep.parse`") for
+   the pattern.
+3. **Exclude it with `# simplecov:disable`.** Reserved for defensive guards that could
+   only be reached by breaking an OS-level invariant. Cover the smallest possible span,
+   state why the code is unreachable, and expect a reviewer to question it. `lib/`
+   currently contains no coverage directives.
+
+   Use the inline form wherever the exclusion is a single line — it applies only to the
+   line it sits on and needs no matching `enable`, which makes it impossible to leave a
+   region accidentally open:
+
+   ```ruby
+   raise 'unreachable' # simplecov:disable defensive guard; only reachable on OOM
+   ```
+
+   The block form covers a span and stays in effect until the matching
+   `# simplecov:enable` (or end of file, if you forget it):
+
+   ```ruby
+   # simplecov:disable branch platform-specific fallback; not reachable on MRI
+   ...
+   # simplecov:enable branch
+   ```
+
+   Always name the narrowest criterion that solves the problem — `line`, `branch`,
+   `method`, or a comma-separated combination — and spell it exactly. A word SimpleCov
+   does not recognize is treated as free-form reason text, which silently widens the
+   directive to all three criteria instead of failing. Write the reason after the
+   criteria, so the required justification lives in the directive itself.
+
+**Coverage is a floor on evidence, not a proof of correctness.** A test written only
+to execute a line, without asserting a meaningful outcome, violates Rule 24 of the
+[`rspec-unit-testing-standards`](.github/skills/rspec-unit-testing-standards/SKILL.md)
+skill and will be rejected in review even though it turns the report green. The
+threshold exists to surface untested behavior, not to be satisfied.
+
+To see exactly what is uncovered:
+
+```bash
+# Names every uncovered line and branch (printed automatically when a full run fails)
+$ LIST_UNCOVERED_DETAIL=true bundle exec rake spec:unit
+
+# Browsable HTML report, also uploaded as a CI artifact when a build fails
+$ open coverage/index.html
+```
+
+This policy applies to `main` only. The `4.x` maintenance branch predates it and is
+not held to these thresholds.
+
 #### Unit tests vs Integration tests
 
 This project uses two types of RSpec tests, organized by directory:

data/Rakefile

@@ -9,19 +9,8 @@ default_tasks = %i[spec:unit spec:integration rubocop]
 default_tasks << :yard if Rake::Task.task_defined?(:yard)
 default_tasks << :build
 
-default_tasks_parallel = %w[spec:unit:parallel spec:integration:parallel rubocop]
-default_tasks_parallel << :yard if Rake::Task.task_defined?(:yard)
-default_tasks_parallel << :build
-
-# Use parallel test execution for MRI where it cuts build times by 30-48%.
-# JRuby and TruffleRuby are slower with parallel_tests because each worker
-# process pays JVM/Truffle startup and warm-up overhead independently,
-# resulting in 18-28% slower builds vs. serial execution.
-parallel_supported = RUBY_ENGINE == 'ruby'
-task default: parallel_supported ? default_tasks_parallel : default_tasks
-
-desc 'Same as default but runs tests in parallel'
-task 'default:parallel': default_tasks_parallel
+desc 'Run all CI tasks (tests, linters, yard, and build)'
+task default: default_tasks
 
 module Rake
   # Overload Rake::Task to add logging

data/git.gemspec

@@ -1,7 +1,19 @@
 # frozen_string_literal: true
 
-$LOAD_PATH.unshift File.expand_path('lib', __dir__)
-require 'git/version'
+# Read the version out of lib/git/version.rb rather than requiring it.
+#
+# The Gemfile uses `gemspec`, so Bundler evaluates this file on every `bundle exec`.
+# `require`ing lib/git/version.rb here would therefore load it before SimpleCov
+# starts in spec_helper, leaving it invisible to the coverage report -- an
+# unmeasured hole in the 100% coverage gate. See the "Test coverage policy" section
+# of CONTRIBUTING.md.
+#
+# release-please rewrites the `VERSION = '...'` assignment in place (see
+# `version-file` in .release-please-config.json), so matching on it stays correct
+# across releases. If the match ever fails, the raise below stops the build
+# immediately rather than letting a nil version reach the specification.
+version = File.read(File.expand_path('lib/git/version.rb', __dir__))[/^\s*VERSION\s*=\s*['"]([^'"]+)['"]/, 1]
+raise 'Could not determine Git::VERSION from lib/git/version.rb' if version.nil?
 
 Gem::Specification.new do |spec|
   spec.author = 'Scott Chacon and others'
@@ -17,7 +29,7 @@ Gem::Specification.new do |spec|
     including branching and merging, object inspection and manipulation, history, patch
     generation and more.
   DESCRIPTION
-  spec.version = Git::VERSION
+  spec.version = version
 
   spec.metadata['homepage_uri'] = spec.homepage
   spec.metadata['source_code_uri'] = spec.homepage
@@ -41,9 +53,9 @@ Gem::Specification.new do |spec|
   spec.add_development_dependency 'rake', '~> 13.3'
   spec.add_development_dependency 'rspec', '~> 3.13'
   spec.add_development_dependency 'rubocop', '~> 1.82'
-  spec.add_development_dependency 'simplecov', '~> 0.22'
+  spec.add_development_dependency 'simplecov', '~> 1.0'
   spec.add_development_dependency 'simplecov-lcov', '~> 0.9'
-  spec.add_development_dependency 'simplecov-rspec', '~> 0.4'
+  spec.add_development_dependency 'simplecov-rspec', '~> 1.1'
 
   if RUBY_ENGINE == 'truffleruby' && Gem::Version.new(RUBY_ENGINE_VERSION) < Gem::Version.new('34.0.0')
     # i18n 1.15+ uses Fiber.[] (Ruby 3.2 Fiber storage) which TruffleRuby < 34.0.0 does not implement

data/lib/git/commands/arguments.rb

@@ -2399,11 +2399,11 @@ module Git
           build_option(args, entry[:name], @option_definitions[entry[:name]], normalized_opts[entry[:name]])
         when :operand
           build_single_positional(args, entry[:name], allocated_positionals)
-        # :nocov: this case should be unreachable
+        # simplecov:disable this case should be unreachable
         else
           raise ArgumentError, "unknown entry kind: #{entry[:kind].inspect}"
         end
-        # :nocov:
+        # simplecov:enable
       end
 
       # Replace the END_OF_OPTIONS_MARKER with the stored `as:` value if any element

data/lib/git/commands/remote/remove.rb

@@ -13,9 +13,9 @@ module Git
       #   remove = Git::Commands::Remote::Remove.new(execution_context)
       #   remove.call('origin')
       #
-      # @example Remove a remote with a name that looks like a flag
-      #   remove = Git::Commands::Remote::Remove.new(execution_context)
-      #   remove.call('-weirdremote')
+      # @note No `end_of_options` boundary: `git remote remove` did not accept a `--`
+      #   separator until git 2.38.0, so a name that looks like a flag (e.g. `-weird`)
+      #   raises `ArgumentError` instead of being sent to git.
       #
       # @note `arguments` block audited against https://git-scm.com/docs/git-remote/2.53.0
       #
@@ -30,8 +30,6 @@ module Git
           literal 'remote'
           literal 'remove'
 
-          end_of_options
-
           operand :name, required: true
         end
 

data/lib/git/commands/remote/rename.rb

@@ -22,6 +22,13 @@ module Git
       #   rename = Git::Commands::Remote::Rename.new(execution_context)
       #   rename.call('origin', 'upstream', no_progress: true)
       #
+      # @note No `end_of_options` boundary despite `:progress` preceding the operands:
+      #   `git remote rename` did not accept a `--` separator until git 2.36.0. Since
+      #   the `old`/`new` operands are still validated (an operand is rejected as
+      #   option-like whenever no `--` boundary exists anywhere in the definition),
+      #   a name that looks like a flag (e.g. `-weird`) raises `ArgumentError` instead
+      #   of ever reaching git, so omitting the boundary introduces no ambiguity.
+      #
       # @note `arguments` block audited against https://git-scm.com/docs/git-remote/2.53.0
       #
       # @see Git::Commands::Remote
@@ -36,8 +43,6 @@ module Git
           literal 'rename'
           flag_option :progress, negatable: true
 
-          end_of_options
-
           operand :old, required: true
           operand :new, required: true
         end

data/lib/git/parsers/diff.rb

@@ -593,11 +593,10 @@ module Git
           # @return [void]
           #
           def apply_file_mode(type, mode)
-            case type
-            when 'new'
+            if type == 'new'
               @current_file[:dst_mode] = mode
               @current_file[:src_path] = nil
-            when 'deleted'
+            else
               @current_file[:src_mode] = mode
               @current_file[:dst_path] = nil
             end

data/lib/git/parsers/grep.rb

@@ -32,7 +32,7 @@ module Git
       def parse(output)
         output.each_line.with_object(Hash.new { |h, k| h[k] = [] }) do |line, hsh|
           filename, line_num, text = line.chomp.split("\0", 3)
-          next unless text && line_num&.match?(/\A\d+\z/)
+          next unless text && line_num.match?(/\A\d+\z/)
 
           hsh[filename] << [line_num.to_i, text]
         end

data/lib/git/version.rb

@@ -4,7 +4,7 @@ module Git
   # The current gem version
   #
   # @return [String] the current gem version
-  VERSION = '5.0.2'
+  VERSION = '5.0.3'
 
   # Represents a git version with major, minor, and patch components
   #

data/tasks/rspec.rake

@@ -1,48 +1,111 @@
 # frozen_string_literal: true
 
-require 'rspec/core/rake_task'
+# Skip parallel test execution when the PARALLEL_TESTS environment variable is set to a falsy value
+def parallel_tests_disabled_via_env?
+  parallel_tests_env = ENV.fetch('PARALLEL_TESTS', 'true').strip.downcase
+  %w[false 0 no off].include?(parallel_tests_env)
+end
+
+# Use parallel test execution for MRI where it cuts build times by 30-48%.
+# JRuby and TruffleRuby are slower with parallel_tests because each worker
+# process pays JVM/Truffle startup and warm-up overhead independently,
+# resulting in 18-28% slower builds vs. serial execution.
+def parallel_tests?
+  RUBY_ENGINE == 'ruby' && !parallel_tests_disabled_via_env?
+end
 
-# Run all specs
-RSpec::Core::RakeTask.new(:spec) do |t|
-  t.pattern = 'spec/**/*_spec.rb'
+# SPEC lets a developer target one or more specific files, e.g.
+# `SPEC=spec/unit/git/version_spec.rb rake spec:unit`, running just those files
+# instead of the whole directory. Both rspec and parallel_rspec recurse a bare
+# directory argument using their own default *_spec.rb pattern, so no explicit
+# --pattern flag is needed for the non-SPEC case.
+#
+# Matches are filtered to dir so that each task only ever runs the specs it owns,
+# no matter how broad SPEC is. This keeps every task self-consistent (a unit spec
+# can never be run by spec:integration under integration's runner and env) and lets
+# a single cross-layer glob drive `rake spec`, e.g.
+# `SPEC=spec/**/git/commands/add_spec.rb rake spec` runs the unit file under
+# spec:unit and the integration file under spec:integration. A SPEC that matches
+# files elsewhere in spec/ but none under dir yields no targets, and the caller
+# skips the task rather than falling back to running all of dir.
+def spec_targets(dir)
+  spec_files = ENV.fetch('SPEC', nil)&.strip
+  return [dir] if spec_files.nil? || spec_files.empty?
+
+  targets = FileList[spec_files].sort
+  raise "SPEC=#{spec_files.inspect} did not match any files" if targets.empty?
+
+  targets.select { |target| target.delete_prefix('./').start_with?(dir) }
 end
 
-# Run only unit specs (mocked, fast)
-RSpec::Core::RakeTask.new('spec:unit') do |t|
-  t.pattern = 'spec/unit/**/*_spec.rb'
-  # On JRuby, use 'documentation' formatter so each test name is printed before
-  # it runs. When the suite hangs, the last printed line identifies the blocking
-  # test. Mirrors the same diagnostic added to spec:integration in PR #1274.
-  t.rspec_opts = '--format documentation' if RUBY_ENGINE == 'jruby'
+# True when targets name exactly one spec file (as opposed to a directory, or
+# several files).
+def single_spec_file?(targets)
+  targets.one? && File.file?(targets.first)
 end
 
-# Run only integration specs (real git, slower)
-RSpec::Core::RakeTask.new('spec:integration') do |t|
-  t.pattern = 'spec/integration/**/*_spec.rb'
-  # On JRuby, override the default Fuubar formatter with 'documentation' so that
-  # each test name is printed before it runs. This makes CI logs useful when the
-  # suite hangs: the last printed test is the one that caused the hang.
-  t.rspec_opts = '--format documentation' if RUBY_ENGINE == 'jruby'
+# The runner for targets: parallel_rspec only when the task runs in parallel and
+# there is more than one spec file to spread across workers.
+#
+# A single file always uses plain rspec. Worker startup is pure overhead with nothing
+# to divide, and parallel_tests sets TEST_ENV_NUMBER in every worker, which makes
+# spec_helper fall back from the documentation formatter to plain dots -- the opposite
+# of what a run narrowed to one file wants.
+def spec_runner(targets, parallel:)
+  parallel && !single_spec_file?(targets) ? 'parallel_rspec' : 'rspec'
 end
 
-# Run all specs in parallel
-desc 'Run all specs in parallel'
-task 'spec:parallel' do
-  sh 'bundle exec parallel_rspec spec/'
+# Run the SPEC-filtered targets for dir, or announce a skip when SPEC selected nothing
+# under dir. Silently doing nothing would read as a passing run of the whole directory.
+def run_spec_task(name, dir, env, parallel:)
+  targets = spec_targets(dir)
+
+  if targets.empty?
+    puts "Skipping #{name}: SPEC=#{ENV.fetch('SPEC', nil)} matched no files under #{dir}"
+  else
+    sh env, 'bundle', 'exec', spec_runner(targets, parallel: parallel), *targets
+  end
 end
 
-# Run only unit specs in parallel
-desc 'Run unit specs in parallel'
-task 'spec:unit:parallel' do
-  sh 'bundle exec parallel_rspec spec/unit/'
+def define_parallel_spec_task(name, dir, env: {})
+  task name do
+    run_spec_task(name, dir, env, parallel: true)
+  end
 end
 
-# Run only integration specs in parallel
-desc 'Run integration specs in parallel'
-task 'spec:integration:parallel' do
-  sh 'bundle exec parallel_rspec spec/integration/'
+def define_serial_spec_task(name, dir, env: {})
+  task name do
+    run_spec_task(name, dir, env, parallel: false)
+  end
 end
 
+# Define a spec task that runs in parallel (via parallel_tests) when parallel_tests? is
+# true, and serially (via a direct `bundle exec rspec` call) otherwise. env, when given,
+# is passed to `sh` and scoped to just the spawned test subprocess(es).
+def define_spec_task(name, dir, desc_text, env: {})
+  desc desc_text
+  parallel_tests? ? define_parallel_spec_task(name, dir, env: env) : define_serial_spec_task(name, dir, env: env)
+end
+
+# Run only unit specs. These run in a few seconds, so parallel_tests startup
+# overhead isn't worth it - always run serially.
+desc 'Run unit RSpec tests (mocked, fast; SPEC=<glob> to run specific files)'
+define_serial_spec_task('spec:unit', 'spec/unit/')
+
+# Run only integration specs. Coverage is forced off (regardless of the COVERAGE
+# env var): integration tests intentionally aren't exhaustive, so tracking or
+# reporting coverage for them would misleadingly suggest that low integration
+# coverage is a problem to fix.
+define_spec_task(
+  'spec:integration', 'spec/integration/',
+  'Run integration RSpec tests (real git, slower; SPEC=<glob> to run specific files)',
+  env: { 'COVERAGE' => 'false' }
+)
+
+# Run all specs, keeping unit and integration output clearly separated
+desc 'Run unit and integration RSpec tests (SPEC=<glob> to run specific files)'
+task spec: %w[spec:unit spec:integration]
+
 CLEAN << 'coverage'
 CLEAN << '.rspec_status'
 CLEAN << 'rspec-report.xml'

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: git
 version: !ruby/object:Gem::Version
-  version: 5.0.2
+  version: 5.0.3
 platform: ruby
 authors:
 - Scott Chacon and others
@@ -169,14 +169,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.22'
+        version: '1.0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.22'
+        version: '1.0'
 - !ruby/object:Gem::Dependency
   name: simplecov-lcov
   requirement: !ruby/object:Gem::Requirement
@@ -197,14 +197,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.4'
+        version: '1.1'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.4'
+        version: '1.1'
 - !ruby/object:Gem::Dependency
   name: irb
   requirement: !ruby/object:Gem::Requirement
@@ -298,6 +298,9 @@ executables: []
 extensions: []
 extra_rdoc_files: []
 files:
+- ".claude/commands/address-copilot-reviews.md"
+- ".claude/settings.json"
+- ".claude/skills"
 - ".commitlintrc.yml"
 - ".dockerignore"
 - ".github/copilot-instructions.md"
@@ -356,6 +359,7 @@ files:
 - ".yardopts"
 - AI_POLICY.md
 - CHANGELOG.md
+- CLAUDE.md
 - CODE_OF_CONDUCT.md
 - CONTRIBUTING.md
 - GOVERNANCE.md
@@ -626,8 +630,8 @@ licenses:
 metadata:
   homepage_uri: http://github.com/ruby-git/ruby-git
   source_code_uri: http://github.com/ruby-git/ruby-git
-  changelog_uri: https://rubydoc.info/gems/git/5.0.2/file/CHANGELOG.md
-  documentation_uri: https://rubydoc.info/gems/git/5.0.2
+  changelog_uri: https://rubydoc.info/gems/git/5.0.3/file/CHANGELOG.md
+  documentation_uri: https://rubydoc.info/gems/git/5.0.3
   rubygems_mfa_required: 'true'
 rdoc_options: []
 require_paths: