@mrciphersmith/keryx 0.2.9

package/src/gdskills/bundled/skills/orchestration/feature-analyzer/orchestrator-prompt.md

@@ -0,0 +1,362 @@
+# Feature Analyzer — Orchestrator Prompt
+
+<!--
+  НАЗНАЧЕНИЕ
+  ==========
+  Этот промпт используется агентом-оркестратором для запуска feature-analyzer
+  как субагента БЕЗ интерактивности. Оркестратор:
+  1. Читает заполненный analysis-request.md (input-файл)
+  2. Парсит параметры из таблиц
+  3. Формирует единый промпт для субагента
+  4. Запускает субагент через Task tool
+  5. Получает результат — путь к сгенерированной документации
+
+  ПОТОК ДАННЫХ
+  ============
+  [Пользователь] → заполняет analysis-request.md
+                    ↓
+  [Оркестратор]  → читает MD → парсит → формирует промпт → Task(subagent)
+                    ↓
+  [Субагент]     → выполняет SKILL.md feature-analyzer автономно
+                    ↓
+  [Результат]    → docs/analysis/<feature>-<date>/
+-->
+
+## Инструкция для оркестратора
+
+### Шаг 1: Прочитать input-файл
+
+Прочитай файл analysis-request с параметрами анализа. Файл может находиться:
+- По пути указанному пользователем
+- В `.metaproject/skills/gdskills/orchestration/feature-analyzer/analysis-request.md` (если пользователь заполнил шаблон)
+
+Извлеки из таблиц следующие значения (парсинг Markdown-таблиц):
+
+```
+SOURCE:
+  local_path     → из таблицы "Source Repository", строка "Local Path"
+  github_repo    → из таблицы "Source Repository", строка "GitHub Repo"
+  branch         → из таблицы "Source Repository", строка "Branch"
+
+TARGET:
+  local_path     → из таблицы "Target Repository", строка "Local Path"
+  github_repo    → из таблицы "Target Repository", строка "GitHub Repo"
+  branch         → из таблицы "Target Repository", строка "Branch"
+
+MODE:
+  mode           → из секции "Analysis Mode", значение после "Mode:"
+  base_branch    → из секции "Analysis Mode", значение после "Base Branch"
+
+FOCUS:
+  description    → из таблицы "Focus", строка "Description"
+  keywords       → из таблицы "Focus", строка "Keywords" (split by comma)
+  directories    → из таблицы "Focus", строка "Directories" (split by comma)
+  scope          → из таблицы "Focus", строка "Scope" (split by comma)
+
+TICKET:
+  url            → из таблицы "Ticket", строка "URL"
+
+AUTOMATION:
+  cache_strategy              → из таблицы "Automation Settings"
+  large_changeset_action      → из таблицы "Automation Settings"
+  large_changeset_max_files   → из таблицы "Automation Settings"
+  timeout_strategy            → из таблицы "Automation Settings"
+  intermediate_review         → из таблицы "Automation Settings"
+  github_mcp_fallback         → из таблицы "Automation Settings"
+
+OUTPUT:
+  base_dir       → из таблицы "Output Configuration"
+  folder_name    → из таблицы "Output Configuration"
+  languages      → из таблицы "Output Configuration" (split by comma)
+```
+
+### Шаг 2: Валидация
+
+Проверь обязательные поля:
+
+```
+ASSERT source.local_path   IS NOT EMPTY  → иначе ABORT("Source path missing")
+ASSERT source.branch       IS NOT EMPTY  → иначе ABORT("Source branch missing")
+ASSERT target.local_path   IS NOT EMPTY  → иначе ABORT("Target path missing")
+ASSERT mode                IN ["A", "B"] → иначе ABORT("Invalid mode")
+IF mode == "A":
+  ASSERT base_branch       IS NOT EMPTY  → иначе ABORT("Mode A requires base_branch")
+```
+
+### Шаг 3: Сформировать промпт для субагента
+
+Подставь извлечённые значения в шаблон ниже и запусти через Task tool.
+
+---
+
+## Шаблон промпта для субагента
+
+```
+You are running the feature-analyzer skill in AUTONOMOUS MODE.
+All interactive checkpoints are pre-resolved. DO NOT ask the user any questions.
+DO NOT stop for confirmation. Execute the full workflow end-to-end.
+
+Load the skill: feature-analyzer (from .metaproject/skills/gdskills/orchestration/feature-analyzer/SKILL.md)
+
+═══════════════════════════════════════════════
+  PRE-RESOLVED CONTEXT (Guard Clause SATISFIED)
+═══════════════════════════════════════════════
+
+SOURCE REPOSITORY:
+  Local path:  {{source.local_path}}
+  GitHub repo: {{source.github_repo}}
+  Branch:      {{source.branch}}
+
+TARGET REPOSITORY:
+  Local path:  {{target.local_path}}
+  GitHub repo: {{target.github_repo}}
+  Branch:      {{target.branch}}
+
+ANALYSIS MODE: {{mode}}
+{{IF mode == "A"}}
+  Base branch: {{base_branch}}
+  → Run Mode A: Changes Analysis (git diff merge-base..HEAD)
+{{ELSE}}
+  → Run Mode B: Current State Analysis (explore entire codebase)
+{{ENDIF}}
+
+{{IF focus.description}}
+FOCUS AREA:
+  Description:  {{focus.description}}
+  Keywords:     {{focus.keywords}}
+  Directories:  {{focus.directories}}
+  Scope:        {{focus.scope}}
+  → Apply focus-based priority boost per SKILL.md rules
+{{ENDIF}}
+
+{{IF ticket.url}}
+TICKET REFERENCE:
+  URL: {{ticket.url}}
+  → Fetch via GitHub MCP and extract business context
+{{ENDIF}}
+
+═══════════════════════════════════════════════
+  AUTOMATION OVERRIDES (All interactions pre-resolved)
+═══════════════════════════════════════════════
+
+1. GUARD CLAUSE CONFIRMATION:
+   → SKIP. Context is pre-validated. Proceed immediately.
+
+2. ANALYSIS MODE SELECTION:
+   → RESOLVED as Mode {{mode}}. Do not ask.
+
+3. BASE BRANCH (Mode A):
+   → RESOLVED as "{{base_branch}}". Do not ask.
+
+4. EXISTING CACHE/REPORTS:
+   → Strategy: {{automation.cache_strategy}}
+   {{IF automation.cache_strategy == "fresh"}}
+     → Ignore all existing analyses. Create new.
+   {{ELIF automation.cache_strategy == "reuse_if_valid"}}
+     → Check cache. If <7 days old AND same SHA → reuse. Else create new.
+   {{ELIF automation.cache_strategy == "refresh"}}
+     → Load previous analysis, analyze only new commits, merge results.
+   {{ELIF automation.cache_strategy == "compare"}}
+     → Create new analysis AND generate diff with existing.
+   {{ENDIF}}
+
+5. LARGE CHANGESET (50+ files):
+   → Strategy: {{automation.large_changeset_action}}
+   → Max files: {{automation.large_changeset_max_files}}
+   {{IF automation.large_changeset_action == "p0_only"}}
+     → Analyze only P0 priority files, max {{automation.large_changeset_max_files}}.
+   {{ELIF automation.large_changeset_action == "directory"}}
+     → Filter to directories: {{automation.large_changeset_filter_paths}}
+   {{ELIF automation.large_changeset_action == "file_types"}}
+     → Filter to extensions: {{automation.large_changeset_filter_extensions}}
+   {{ELIF automation.large_changeset_action == "commits"}}
+     → Analyze first {{automation.large_changeset_max_commits}} commits only.
+   {{ELIF automation.large_changeset_action == "abort"}}
+     → Return error: "Changeset too large, aborting per automation config."
+   {{ENDIF}}
+
+6. TIMEOUT (approaching 30 min):
+   → Strategy: {{automation.timeout_strategy}}
+   {{IF automation.timeout_strategy == "partial"}}
+     → Generate report with whatever findings are available at that point.
+   {{ELIF automation.timeout_strategy == "extend"}}
+     → Continue for 15 more minutes, then force partial.
+   {{ELIF automation.timeout_strategy == "prioritize_p0"}}
+     → Drop all non-P0 files and finish only P0 analysis.
+   {{ELIF automation.timeout_strategy == "abort"}}
+     → Return error: "Analysis timed out, aborting per automation config."
+   {{ENDIF}}
+
+7. INTERMEDIATE REVIEW:
+   → Strategy: {{automation.intermediate_review}}
+   {{IF automation.intermediate_review == "skip"}}
+     → Do NOT show intermediate summary. Proceed directly to documentation.
+   {{ELIF automation.intermediate_review == "log_only"}}
+     → Log intermediate summary in output but do not wait for confirmation.
+   {{ELIF automation.intermediate_review == "include_in_report"}}
+     → Include intermediate findings as a section in the final report.
+   {{ENDIF}}
+
+8. GITHUB MCP UNAVAILABLE:
+   → Fallback: {{automation.github_mcp_fallback}}
+   {{IF automation.github_mcp_fallback == "git_only"}}
+     → Proceed with local git data only. No GitHub API calls.
+   {{ELIF automation.github_mcp_fallback == "abort"}}
+     → Return error: "GitHub MCP unavailable, aborting per automation config."
+   {{ENDIF}}
+
+═══════════════════════════════════════════════
+  OUTPUT CONFIGURATION
+═══════════════════════════════════════════════
+
+Base directory: {{output.base_dir | default("<DOCS_ROOT>/analysis")}}
+Folder name:    {{output.folder_name | default("auto-generated per SKILL.md rules")}}
+Languages:      {{output.languages | default("en, ru, ai")}}
+Include metrics: {{output.include_metrics | default("true")}}
+
+═══════════════════════════════════════════════
+  EXECUTION INSTRUCTIONS
+═══════════════════════════════════════════════
+
+1. Load feature-analyzer SKILL.md
+2. Skip directly to Step 1 (context is pre-provided above)
+3. Follow the workflow for Mode {{mode}} from SKILL.md
+4. At every interaction checkpoint (marked above), use the pre-resolved strategy
+5. Generate all output documents per SKILL.md structure
+6. Return the following in your final message:
+   - Path to generated analysis directory
+   - Summary of key findings (3-5 bullet points)
+   - Complexity score and risk level
+   - List of generated files
+
+DO NOT ask questions. DO NOT stop for user input. Run to completion.
+```
+
+---
+
+## Пример вызова оркестратором
+
+### Input: Заполненный analysis-request.md
+
+```markdown
+## Source Repository *
+| Field | Value |
+|-------|-------|
+| Local Path * | `/Users/dev/<PROJECT>` |
+| GitHub Repo | `<ORG>/<PROJECT>` |
+| Branch * | `feature/pipeline-variables` |
+
+## Target Repository *
+| Field | Value |
+|-------|-------|
+| Local Path * | `/Users/dev/<PROJECT>` |
+| Branch | `main` |
+
+## Analysis Mode *
+- **Mode**: `A`
+- **Base Branch**: `main`
+
+## Focus
+| Field | Value |
+|-------|-------|
+| Description | `variables in pipelines` |
+| Keywords | `variable, pipeline, param` |
+| Directories | `src/pipelines, src/models` |
+
+## Automation Settings *
+| Setting | Value |
+|---------|-------|
+| skip_confirmation | `true` |
+| cache_strategy | `fresh` |
+| large_changeset_action | `p0_only` |
+| large_changeset_max_files | `7` |
+| timeout_strategy | `partial` |
+| intermediate_review | `include_in_report` |
+| github_mcp_fallback | `git_only` |
+```
+
+### Output: Сгенерированный промпт (после подстановки)
+
+```
+You are running the feature-analyzer skill in AUTONOMOUS MODE.
+All interactive checkpoints are pre-resolved. DO NOT ask the user any questions.
+...
+
+SOURCE REPOSITORY:
+  Local path:  /Users/dev/<PROJECT>
+  GitHub repo: <ORG>/<PROJECT>
+  Branch:      feature/pipeline-variables
+
+TARGET REPOSITORY:
+  Local path:  /Users/dev/<PROJECT>
+  Branch:      main
+
+ANALYSIS MODE: A
+  Base branch: main
+  → Run Mode A: Changes Analysis (git diff merge-base..HEAD)
+
+FOCUS AREA:
+  Description:  variables in pipelines
+  Keywords:     variable, pipeline, param
+  Directories:  src/pipelines, src/models
+  → Apply focus-based priority boost per SKILL.md rules
+...
+```
+
+### Вызов через Task tool
+
+```javascript
+Task({
+  description: "Feature analysis: pipeline-variables",
+  subagent_type: "general",
+  prompt: "<сгенерированный промпт выше>"
+})
+```
+
+---
+
+## Карта устранённых интерактивных точек
+
+| # | Interaction Point | Где в SKILL.md | Как устраняется |
+|---|-------------------|---------------|-----------------|
+| 1 | Guard Clause Confirmation | L22-54 | `skip_confirmation: true` + pre-provided context |
+| 2 | Mode Selection | L576-589 | `analysis_mode.mode: "A"/"B"` |
+| 3 | Base Branch Query | L650-670 | `analysis_mode.base_branch: "main"` |
+| 4 | Existing Cache Dialog | L489-510 | `cache_strategy: "fresh"/"reuse_if_valid"/...` |
+| 5 | Large Changeset Dialog | L442-457 | `large_changeset_strategy.action: "p0_only"/...` |
+| 6 | Timeout Dialog | L427-431 | `timeout_strategy: "partial"/"extend"/...` |
+| 7 | Intermediate Review | L1009-1046 | `intermediate_review: "skip"/"log_only"/...` |
+
+---
+
+## Рекомендуемые automation-профили
+
+### Profile: Quick (быстрый анализ, минимум взаимодействия)
+```
+cache_strategy:         reuse_if_valid
+large_changeset_action: p0_only
+large_changeset_max_files: 5
+timeout_strategy:       partial
+intermediate_review:    skip
+github_mcp_fallback:    git_only
+```
+
+### Profile: Thorough (полный анализ, максимум данных)
+```
+cache_strategy:         fresh
+large_changeset_action: p0_only
+large_changeset_max_files: 10
+timeout_strategy:       extend
+intermediate_review:    include_in_report
+github_mcp_fallback:    git_only
+```
+
+### Profile: CI/CD (для автоматических пайплайнов)
+```
+cache_strategy:         fresh
+large_changeset_action: p0_only
+large_changeset_max_files: 7
+timeout_strategy:       partial
+intermediate_review:    skip
+github_mcp_fallback:    git_only
+```

package/src/gdskills/bundled/skills/orchestration/feature-analyzer/output-contract.schema.json

@@ -0,0 +1,87 @@
+{
+  "$schema": "https://json-schema.org/draft/2020-12/schema",
+  "$id": "https://keryx.local/schemas/feature-analyzer/output-contract.schema.json",
+  "title": "Feature Analyzer Output Contract",
+  "description": "Structured analysis output. Delivered as a Markdown report following SKILL.md output format.",
+  "type": "object",
+  "additionalProperties": false,
+  "required": ["mode", "source", "target", "summary", "files_analyzed"],
+  "properties": {
+    "mode": {
+      "type": "string",
+      "enum": ["A", "B"],
+      "description": "Analysis mode: A = branch changes, B = current state"
+    },
+    "source": {
+      "type": "object",
+      "required": ["path", "branch"],
+      "properties": {
+        "path": { "type": "string" },
+        "branch": { "type": "string" },
+        "merge_base": { "type": "string" }
+      }
+    },
+    "target": {
+      "type": "object",
+      "required": ["path"],
+      "properties": {
+        "path": { "type": "string" },
+        "branch": { "type": "string" }
+      }
+    },
+    "focus": {
+      "type": "string",
+      "description": "Focus area description as provided by user"
+    },
+    "summary": {
+      "type": "string",
+      "description": "Executive summary of analysis findings",
+      "minLength": 1
+    },
+    "files_analyzed": {
+      "type": "object",
+      "properties": {
+        "p0": {
+          "type": "array",
+          "description": "Must-change files",
+          "items": {
+            "type": "object",
+            "required": ["path", "reason"],
+            "properties": {
+              "path": { "type": "string" },
+              "reason": { "type": "string" }
+            }
+          }
+        },
+        "p1": {
+          "type": "array",
+          "description": "Likely-change files",
+          "items": { "type": "object", "required": ["path", "reason"], "properties": { "path": { "type": "string" }, "reason": { "type": "string" } } }
+        },
+        "p2": {
+          "type": "array",
+          "description": "May-need-update files (tests, stories, docs)",
+          "items": { "type": "object", "required": ["path", "reason"], "properties": { "path": { "type": "string" }, "reason": { "type": "string" } } }
+        }
+      }
+    },
+    "api_contracts": {
+      "type": "array",
+      "description": "API endpoints or contracts identified",
+      "items": {
+        "type": "object",
+        "properties": {
+          "name": { "type": "string" },
+          "type": { "type": "string" },
+          "change": { "type": "string", "enum": ["added", "modified", "removed", "unchanged"] },
+          "details": { "type": "string" }
+        }
+      }
+    },
+    "recommendations": {
+      "type": "array",
+      "description": "Recommended implementation steps for the target repository",
+      "items": { "type": "string" }
+    }
+  }
+}

package/src/gdskills/bundled/skills/orchestration/feature-dev/SKILL.codex.md

@@ -0,0 +1,105 @@
+---
+name: feature-dev
+description: "7-phase feature development workflow: requirements → design → implement → test → review → fix → PR. Full cycle from idea or GitHub issue to merge-ready PR. Confirms requirements and design with user before autonomous execution."
+triggers:
+  - "/feature-dev"
+  - "Develop feature"
+  - "Build feature"
+  - "Implement feature"
+  - "Feature from scratch"
+metadata:
+  author: "MrCipherSmith"
+  version: "1.0.0"
+  category: "workflow"
+license: "MIT"
+compatibility: "cursor,codex,zed,opencode,claude"
+---
+
+# Feature Development (7-Phase)
+
+End-to-end feature development workflow from idea to merge-ready PR.
+
+## Arguments
+
+- `/feature-dev <description>` — start from a text description
+- `/feature-dev #<issue>` — start from a GitHub issue
+- `/feature-dev --resume` — resume interrupted feature-dev (checks for existing worktree/branch)
+
+## 7-Phase Architecture
+
+### Phase 1: REQUIREMENTS
+1. Parse input (description or GitHub issue via `gh issue view`)
+2. Clarify ambiguities — ask the user up to 3 questions max
+3. Produce a brief spec:
+   - **What**: feature description in 2-3 sentences
+   - **Why**: user value / business reason
+   - **Scope**: what's in, what's explicitly out
+   - **Acceptance criteria**: testable bullet points
+4. **Get user confirmation before proceeding**
+
+### Phase 2: DESIGN
+1. Research the codebase:
+   - Find related modules via search tools
+   - Read neighboring implementations for patterns
+   - Check existing tests for testing conventions
+2. Produce implementation plan:
+   - Files to create/modify (with brief description of changes)
+   - Dependencies or packages needed
+   - Data model changes if any
+   - Estimated complexity: S / M / L
+3. **Get user confirmation on the plan**
+
+### Phase 3: PREPARE
+1. Create a feature branch: `git checkout -b feat/<name>`
+2. If large feature, consider `git worktree add` for isolation
+3. Install any new dependencies
+
+### Phase 4: IMPLEMENT
+1. Implement changes file by file, following the plan
+2. Follow existing code patterns discovered in Phase 2
+3. After each logical chunk, run available checks:
+   - Lint: `npm run lint` or equivalent
+   - Type-check: `npx tsc --noEmit` or equivalent
+   - Fix issues immediately before moving on
+
+### Phase 5: TEST
+1. Write tests matching the project's testing patterns
+2. Unit tests for new functions/modules
+3. Integration tests for API/data flow changes
+4. Run full test suite
+5. Fix failing tests (max 3 attempts per test)
+
+### Phase 6: REVIEW (Self)
+1. Launch `code-review` skill on own changes (if available)
+2. Or run a focused self-review:
+   - `git diff main...HEAD` — review the full diff
+   - Check for: TODOs left behind, console.logs, hardcoded values
+   - Verify all acceptance criteria from Phase 1
+3. Fix any findings (max 2 review-fix cycles)
+
+### Phase 7: DELIVER
+1. Final checks: lint + type-check + tests all pass
+2. Commit with conventional message: `feat(<scope>): <description>`
+3. Push branch
+4. Create PR:
+   - Link to issue if applicable
+   - Include acceptance criteria as checklist
+   - Add test plan
+5. Report PR URL to user
+
+## Status Updates
+
+At each phase transition, report progress:
+```
+✅ Phase 1: Requirements confirmed
+🔄 Phase 2: Designing implementation...
+```
+
+## Rules
+
+- ALWAYS get user confirmation after Phase 1 (requirements) and Phase 2 (design)
+- Phases 4-6 are autonomous — no user interaction needed
+- If stuck for >3 attempts on any step, report the blocker and ask user
+- NEVER skip Phase 5 (testing) even if user says "skip tests"
+- NEVER commit broken code (lint/type-check must pass)
+- Keep commits atomic: one commit per logical change, not one giant commit

package/src/gdskills/bundled/skills/orchestration/feature-dev/SKILL.cursor.md

@@ -0,0 +1,105 @@
+---
+name: feature-dev
+description: "7-phase feature development workflow: requirements → design → implement → test → review → fix → PR. Full cycle from idea or GitHub issue to merge-ready PR. Confirms requirements and design with user before autonomous execution."
+triggers:
+  - "/feature-dev"
+  - "Develop feature"
+  - "Build feature"
+  - "Implement feature"
+  - "Feature from scratch"
+metadata:
+  author: "MrCipherSmith"
+  version: "1.0.0"
+  category: "workflow"
+license: "MIT"
+compatibility: "cursor,codex,zed,opencode,claude"
+---
+
+# Feature Development (7-Phase)
+
+End-to-end feature development workflow from idea to merge-ready PR.
+
+## Arguments
+
+- `/feature-dev <description>` — start from a text description
+- `/feature-dev #<issue>` — start from a GitHub issue
+- `/feature-dev --resume` — resume interrupted feature-dev (checks for existing worktree/branch)
+
+## 7-Phase Architecture
+
+### Phase 1: REQUIREMENTS
+1. Parse input (description or GitHub issue via `gh issue view`)
+2. Clarify ambiguities — ask the user up to 3 questions max
+3. Produce a brief spec:
+   - **What**: feature description in 2-3 sentences
+   - **Why**: user value / business reason
+   - **Scope**: what's in, what's explicitly out
+   - **Acceptance criteria**: testable bullet points
+4. **Get user confirmation before proceeding**
+
+### Phase 2: DESIGN
+1. Research the codebase:
+   - Find related modules via search tools
+   - Read neighboring implementations for patterns
+   - Check existing tests for testing conventions
+2. Produce implementation plan:
+   - Files to create/modify (with brief description of changes)
+   - Dependencies or packages needed
+   - Data model changes if any
+   - Estimated complexity: S / M / L
+3. **Get user confirmation on the plan**
+
+### Phase 3: PREPARE
+1. Create a feature branch: `git checkout -b feat/<name>`
+2. If large feature, consider `git worktree add` for isolation
+3. Install any new dependencies
+
+### Phase 4: IMPLEMENT
+1. Implement changes file by file, following the plan
+2. Follow existing code patterns discovered in Phase 2
+3. After each logical chunk, run available checks:
+   - Lint: `npm run lint` or equivalent
+   - Type-check: `npx tsc --noEmit` or equivalent
+   - Fix issues immediately before moving on
+
+### Phase 5: TEST
+1. Write tests matching the project's testing patterns
+2. Unit tests for new functions/modules
+3. Integration tests for API/data flow changes
+4. Run full test suite
+5. Fix failing tests (max 3 attempts per test)
+
+### Phase 6: REVIEW (Self)
+1. Launch `code-review` skill on own changes (if available)
+2. Or run a focused self-review:
+   - `git diff main...HEAD` — review the full diff
+   - Check for: TODOs left behind, console.logs, hardcoded values
+   - Verify all acceptance criteria from Phase 1
+3. Fix any findings (max 2 review-fix cycles)
+
+### Phase 7: DELIVER
+1. Final checks: lint + type-check + tests all pass
+2. Commit with conventional message: `feat(<scope>): <description>`
+3. Push branch
+4. Create PR:
+   - Link to issue if applicable
+   - Include acceptance criteria as checklist
+   - Add test plan
+5. Report PR URL to user
+
+## Status Updates
+
+At each phase transition, report progress:
+```
+✅ Phase 1: Requirements confirmed
+🔄 Phase 2: Designing implementation...
+```
+
+## Rules
+
+- ALWAYS get user confirmation after Phase 1 (requirements) and Phase 2 (design)
+- Phases 4-6 are autonomous — no user interaction needed
+- If stuck for >3 attempts on any step, report the blocker and ask user
+- NEVER skip Phase 5 (testing) even if user says "skip tests"
+- NEVER commit broken code (lint/type-check must pass)
+- Keep commits atomic: one commit per logical change, not one giant commit