cloudnative-co · okash1n · Jun 12, 2026 · Jun 12, 2026 · chatgpt-codex-connector · Jun 12, 2026
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -4,6 +4,21 @@ All notable changes to this project will be documented in this file.
 
 The format is based on [Keep a Changelog](https://keepachangelog.com/), and this project adheres to [Semantic Versioning](https://semver.org/).
 
+## [0.65.0] - 2026-06-12
+
+LLM 性能監査（#121）P3 の第 2 弾（#115）。commands の低優先掃除（ネイティブ重複・呪文・ハードコードの除去）。
+
+### Removed
+- **update-codemaps を削除（#115）**: 静的コードマップの常設維持は旧モデルの探索能力不足への補償で、現行モデルは必要時の直接探索の方が常に新鮮。恣意的な「30% 変更で承認」ゲートと「Use TypeScript/Node.js」指定も同時に廃止。agents/doc-updater の codemap 言及を一般化。doc-blocker の codemaps/ allowlist は既存ユーザーの生成物保護のため残置。既存配備分は update の retired 掃除で除去（コマンド数 21 → 20）
+
+### Changed
+- **checkpoint を git マイルストーン記録に縮小（#115）**: セッション内ロールバックはネイティブ checkpoint/rewind（Esc Esc / /rewind)に委譲することを明記し、本コマンドはセッション横断のマイルストーン記録専用に。記録していないデータ（過去のテスト成績・カバレッジ）との比較を要求していた verify の矛盾を解消（`git diff --stat` + 現在のテスト結果のみ報告）
+- **research のマジックワード呪文を削除（#115）**: 「"deeply examine" 等の語句を使え」という旧世代プロンプティングの迷信を削除（精読の強制自体は維持）
+- **update-kit のツール選択 babysitting を削除（#115）**: 「cat ではなく Read ツールを使え」の注記を削除（conf 編集ガードレールは維持）
+- **test-coverage のスタック固定を一般化（#115）**: npm/pnpm・coverage-summary.json 固定をテストツール検出 + 例示に変更（80% 閾値はキット規約として維持）
+- **e2e / e2e-runner の CI 例からバージョンピンを除去（#115）**: actions@v4 / Node 22 の固定例示が「最新を使え」という指示文と自己矛盾していたため、「生成時に最新メジャーと Active LTS を確認、既存 .github/workflows/ があればそれを基準に」へ変更
+- **spec-kit-init の陳腐化前提を更新（#115）**: 「Spec Kit v0.10 が出たら再評価」の未来予告を v0.10.0 出荷済みの現状記述（git extension は opt-in、--no-git 廃止）に置換。「4 steps」→「5 steps」の不整合も修正
+
 ## [0.64.0] - 2026-06-12
 
 LLM 性能監査（#121）P3 の第 1 弾（#114）。commands の旧モデル向け儀式（固定手順の強制）を成果指向に書き換えた。

diff --git a/README.en.md b/README.en.md
@@ -52,7 +52,7 @@ Claude Code Starter Kit bootstraps a consistent, high-quality Claude Code enviro
 - **3 profiles**: Minimal, Standard (recommended), Full
 - **9 agents**: planner, architect, tdd-guide, code-reviewer, security-reviewer, build-error-resolver, e2e-runner, refactor-cleaner, doc-updater
 - **10 rules**: coding-style, git-workflow, hooks, patterns, performance, security, testing, agents, anti-patterns, permissions-guide
-- **21 slash commands**: /plan, /tdd, /build-fix, /e2e, /verify, /research, /web-article, /oss-analyze, /web-source-review, /handover, /update-kit, and more
+- **20 slash commands**: /plan, /tdd, /build-fix, /e2e, /verify, /research, /web-article, /oss-analyze, /web-source-review, /handover, /update-kit, and more
 - **12 skill modules**: backend-patterns, frontend-patterns, security-review, tdd-workflow, prompt-patterns, and more
 - **11 optional hooks/settings**: safety net (cc-safety-net), auto update, web content update, tmux reminder, doc blocker, Prettier or Biome formatting, PR creation log, pre-compact snapshot (opt-in), statusline, doc size guard, feature recommendation
 - **14 plugins** from multiple marketplaces: security-guidance, commit-commands, pr-review-toolkit, feature-dev, code-review, claude-md-management, superpowers, code-simplifier, document-skills, example-skills, typescript-lsp, gopls-lsp, pyright-lsp, rust-analyzer-lsp
@@ -393,7 +393,7 @@ claude-code-starter-kit/
 │   └── ja/                 # Japanese templates & strings
 ├── agents/                 # Agent definitions (9 files)
 ├── rules/                  # Rule files (10 files)
-├── commands/               # Slash commands (21 files)
+├── commands/               # Slash commands (20 files)
 └── skills/                 # Skill modules (12 dirs)
 ```
 

diff --git a/README.md b/README.md
@@ -432,7 +432,7 @@ cd claude-code-starter-kit
 | **refactor-cleaner** | コード整理 | 不要なコードの削除・整理 |
 | **doc-updater** | ドキュメント更新 | README やドキュメントの更新 |
 
-### 💬 スラッシュコマンド（21個）
+### 💬 スラッシュコマンド（20個）
 
 スラッシュコマンドは、Claude に **ワンタッチで指示を出すショートカット** です。
 Claude Code のチャットで `/` に続けて入力します。
@@ -1104,7 +1104,7 @@ claude-code-starter-kit/
 │   └── ja/                 # 日本語
 ├── agents/                 # AI エージェント定義（9種）
 ├── rules/                  # コーディングルール（10種）
-├── commands/               # スラッシュコマンド（21個）
+├── commands/               # スラッシュコマンド（20個）
 └── skills/                 # スキルモジュール（12個）
 ```
 

diff --git a/agents/doc-updater.md b/agents/doc-updater.md
@@ -1,6 +1,6 @@
 ---
 name: doc-updater
-description: Documentation maintenance specialist. Use when code changes require README, guide, codemap, runbook, or changelog updates.
+description: Documentation maintenance specialist. Use when code changes require README, guide, runbook, or changelog updates.
 tools: Read, Write, Edit, Bash, Grep, Glob
 model: haiku
 ---
@@ -21,7 +21,7 @@ You keep documentation aligned with the repository. Update only documents that a
 
 - Do not invent architecture not visible in the repo.
 - Do not create new docs when a targeted edit is enough.
-- Keep generated codemaps or indexes deterministic.
+- Keep generated docs or indexes (if the project maintains any) deterministic.
 - Mention any documentation area that looks stale but is outside scope.
 
 ## Output

diff --git a/agents/e2e-runner.md b/agents/e2e-runner.md
@@ -26,22 +26,17 @@ You design and run end-to-end tests that protect real user journeys. Prefer a sm
 
 ## CI Example
 
-Use current action versions and an active Node LTS:
+Use the latest major action versions and the active Node LTS at
+generation time (verify rather than recalling pinned versions). The
+step sequence is:
 
 ```yaml
-- uses: actions/checkout@v4
-- uses: actions/setup-node@v4
-  with:
-    node-version: 22
-    cache: npm
+- checkout
+- setup-node (active LTS, cache npm)
 - run: npm ci
 - run: npx playwright install --with-deps
 - run: npm run test:e2e
-- uses: actions/upload-artifact@v4
-  if: failure()
-  with:
-    name: playwright-artifacts
-    path: test-results/
+- upload-artifact (on failure: test-results/)
 ```
 
 ## Output

diff --git a/commands/checkpoint.md b/commands/checkpoint.md
@@ -1,18 +1,18 @@
 # Checkpoint Command
 
-Create or verify a checkpoint in your workflow.
+Record cross-session git milestones. In-session rollback is handled by
+Claude Code's native checkpoint/rewind (Esc Esc / /rewind) — use this
+command only for multi-day milestones you want recorded in git.
 
 ## Usage
 
 `/checkpoint [create|verify|list] [name]`
 
 ## Create Checkpoint
 
-When creating a checkpoint:
-
 1. Run `/verify quick` to ensure current state is clean
-2. Create a git stash or commit with checkpoint name
-3. Log checkpoint to `.claude/checkpoints.log`:
+2. Create a git commit or tag with the checkpoint name
+3. Log the checkpoint to `.claude/checkpoints.log`:
 
 ```bash
 echo "$(date +%Y-%m-%d-%H:%M) | $CHECKPOINT_NAME | $(git rev-parse --short HEAD)" >> .claude/checkpoints.log
@@ -22,53 +22,20 @@ echo "$(date +%Y-%m-%d-%H:%M) | $CHECKPOINT_NAME | $(git rev-parse --short HEAD)
 
 ## Verify Checkpoint
 
-When verifying against a checkpoint:
-
-1. Read checkpoint from log
-2. Compare current state to checkpoint:
-   - Files added since checkpoint
-   - Files modified since checkpoint
-   - Test pass rate now vs then
-   - Coverage now vs then
-
-3. Report:
-```
-CHECKPOINT COMPARISON: $NAME
-============================
-Files changed: X
-Tests: +Y passed / -Z failed
-Coverage: +X% / -Y%
-Build: [PASS/FAIL]
-```
+1. Read the checkpoint SHA from the log
+2. Show what changed since then: `git diff --stat <sha>..HEAD`
+3. Run the test suite once and report the **current** status
+   (the log records only date/name/SHA — there is no historical
+   test/coverage data to compare against)
 
 ## List Checkpoints
 
-Show all checkpoints with:
-- Name
-- Timestamp
-- Git SHA
-- Status (current, behind, ahead)
-
-## Workflow
-
-Typical checkpoint flow:
-
-```
-[Start] --> /checkpoint create "feature-start"
-   |
-[Implement] --> /checkpoint create "core-done"
-   |
-[Test] --> /checkpoint verify "core-done"
-   |
-[Refactor] --> /checkpoint create "refactor-done"
-   |
-[PR] --> /checkpoint verify "feature-start"
-```
+Show all checkpoints with name, timestamp, git SHA, and whether HEAD is
+at, ahead of, or behind each.
 
 ## Arguments
 
 $ARGUMENTS:
-- `create <name>` - Create named checkpoint
-- `verify <name>` - Verify against named checkpoint
-- `list` - Show all checkpoints
-- `clear` - Remove old checkpoints (keeps last 5)
+- `create <name>` - Record a milestone
+- `verify <name>` - Diff + current test status against a milestone
+- `list` - Show recorded milestones
diff --git a/commands/e2e.md b/commands/e2e.md
@@ -19,15 +19,10 @@ Use the `e2e-runner` agent and the repository's existing browser test setup.
 
 ## CI Baseline
 
-When creating GitHub Actions examples, use current actions and active Node LTS:
-
-```yaml
-- uses: actions/checkout@v4
-- uses: actions/setup-node@v4
-  with:
-    node-version: 22
-- uses: actions/upload-artifact@v4
-```
+When creating GitHub Actions examples, check and use the latest major
+versions of actions and the active Node LTS at generation time; if the
+repository already has `.github/workflows/`, treat those as the current
+baseline instead of hardcoding versions.
 
 ## Output
 

diff --git a/commands/research.md b/commands/research.md
@@ -10,7 +10,7 @@ You are entering **Research Phase**. Your goal is to deeply understand the relev
 
 1. **Identify scope**: Based on the user's request, determine which parts of the codebase are relevant.
 
-2. **Deep read**: Read all relevant files thoroughly. Use words like "deeply examine", "trace the flow", "understand the intricacies" in your analysis. Don't skim - understand the actual logic, edge cases, and design decisions.
+2. **Deep read**: Read all relevant files thoroughly. Don't skim - understand the actual logic, edge cases, and design decisions.
 
 3. **Map dependencies**: Trace imports, function calls, and data flow across files. Document the dependency graph.
 

diff --git a/commands/spec-kit-init.md b/commands/spec-kit-init.md
@@ -14,7 +14,7 @@ cycle; `uv tool upgrade specify-cli` is how you keep Spec Kit current.
 ## Instructions
 
 You are entering **Spec Kit Initialization**. Walk the user through the
-4 steps below in order. Do not execute heavy operations without
+5 steps below in order. Do not execute heavy operations without
 confirmation; the user is in control.
 
 ### Step 1 — Verify prerequisites
@@ -58,6 +58,10 @@ specify init . --integration claude --script sh
 specify init --here --integration claude --script sh --force
 ```
 
+If the user wants Spec Kit's feature-branch automation, run
+`specify extension add git` after init (the git extension is opt-in
+since v0.10.0).
+
 After init, the project will have:
 
 - `.claude/skills/speckit-*` — 14 SKILL.md files (project-local, namespaced)
@@ -97,7 +101,7 @@ explicit go-ahead — it executes the entire task list and can be expensive.
 ## Notes & gotchas
 
 - Spec Kit's command separator is **hyphen, not dot**: `/speckit-plan`, NOT `/speckit.plan`. The dot form was retired before v0.8.0.
-- `specify init` creates a git branch unless `--no-git` is passed; Spec Kit v0.10 will make the `git` extension opt-in. Re-evaluate this command when v0.10 ships.
+- Since Spec Kit v0.10.0 (2026-06), the git extension is opt-in and the `--no-git` flag has been removed; `specify init` does not create a git branch by default. Run `specify extension add git` after init if you want Spec Kit's branch automation.
 - `specify init` writes only into `<project>/`. It does NOT touch `~/.claude/` (verified during the StarterKit evaluation in `docs/spec-kit-evaluation.md`).
 - Spec Kit's bash scripts (`.specify/scripts/bash/*.sh`) do NOT use `rm -rf`, `curl`, `wget`, `sudo`, or `git push --force`. They are compatible with the StarterKit's deny rules.
 - Long-term maintenance of `specify-cli` is delegated to `uv tool upgrade`; the StarterKit does not bundle or pin a Spec Kit version.

diff --git a/commands/test-coverage.md b/commands/test-coverage.md
@@ -2,9 +2,11 @@
 
 Analyze test coverage and generate missing tests:
 
-1. Run tests with coverage: npm test --coverage or pnpm test --coverage
+1. Identify the project's test tooling and run tests with coverage
+   (e.g. npm test --coverage, pytest --cov, go test -cover)
 
-2. Analyze coverage report (coverage/coverage-summary.json)
+2. Analyze the coverage report your test runner produces
+   (e.g. coverage/coverage-summary.json for JS)
 
 3. Identify files below 80% coverage threshold
 
@@ -23,5 +25,5 @@ Analyze test coverage and generate missing tests:
 Focus on:
 - Happy path scenarios
 - Error handling
-- Edge cases (null, undefined, empty)
+- Edge cases (null / missing / empty values)
 - Boundary conditions
diff --git a/commands/update-codemaps.md b/commands/update-codemaps.md
diff --git a/commands/update-kit.md b/commands/update-kit.md
@@ -35,7 +35,7 @@ Before running the update:
 
 After a successful update (or if already up to date), check for pending feature recommendations.
 
-**Note**: Always use the Read tool (not `cat`) to read files, and the Edit tool to modify `~/.claude-starter-kit.conf`. Skip any feature named `feature-recommendation` if it appears in the pending list (self-referential — should not happen, but guard against it).
+**Note**: Skip any feature named `feature-recommendation` if it appears in the pending list (self-referential — should not happen, but guard against it).
 
 #### Step 1: Check for pending features
 

diff --git a/skills/eval-harness/SKILL.md b/skills/eval-harness/SKILL.md
@@ -34,7 +34,7 @@ Expected Output: Description of expected result
 Ensure changes don't break existing functionality:
 ```markdown
 [REGRESSION EVAL: feature-name]
-Baseline: SHA or checkpoint name
+Baseline: git SHA (or a /checkpoint milestone name)
 Tests:
   - existing-test-1: PASS/FAIL
   - existing-test-2: PASS/FAIL

diff --git a/tests/unit/test-agent-definitions.sh b/tests/unit/test-agent-definitions.sh
@@ -44,9 +44,12 @@
 }
 
 {
-  test_name="agents: e2e and security guidance use current baselines"
-  if grep -q 'actions/upload-artifact@v4' "$PROJECT_DIR/agents/e2e-runner.md" \
-    && grep -q 'node-version: 22' "$PROJECT_DIR/agents/e2e-runner.md" \
+  test_name="agents: e2e and security guidance avoid stale pinned baselines"
+  if ! grep -q 'actions/checkout@v' "$PROJECT_DIR/agents/e2e-runner.md" \
+    && ! grep -q 'node-version:' "$PROJECT_DIR/agents/e2e-runner.md" \
+    && ! grep -q 'actions/checkout@v' "$PROJECT_DIR/commands/e2e.md" \
+    && ! grep -q 'node-version:' "$PROJECT_DIR/commands/e2e.md" \
+    && grep -q 'active Node LTS' "$PROJECT_DIR/agents/e2e-runner.md" \
     && grep -q 'context.tracing.start' "$PROJECT_DIR/agents/e2e-runner.md" \
     && grep -q 'OWASP Top 10 2021' "$PROJECT_DIR/agents/security-reviewer.md" \
     && grep -q 'Server-Side Request Forgery' "$PROJECT_DIR/agents/security-reviewer.md"; then