feat: add release preparation documentation and tooling (Phase 8) (#61)

## Summary

Completes Phase 8: Release Preparation for Issue #15. This PR adds all
necessary documentation, tooling updates, and test fixes for the first
release.

### Changes

**Documentation:**
- Updated CHANGELOG.md with comprehensive release notes for v1.0.0
- Updated README.md with CI badges, quick start guide, and headless mode
docs
- Added `docs/playbooks/release.md` - complete release checklist and
process
- Added `docs/playbooks/github-setup.md` - secrets and branch protection
guide

**Tooling:**
- Updated ADR-0031 (PR Validation Automation) from proposed to accepted
- Updated Stryker.NET from 4.4.2 to 4.11.0 with new config schema
- Set realistic mutation testing thresholds (baseline: 14.44%)
- Added 'unlist' to project spelling dictionary

**CI/CD:**
- Fixed Dependabot config to use `build` prefix (valid conventional
commit type)
- Previous `deps` prefix was invalid, causing commitlint failures

**Test Isolation:**
- Refactored LogCapture to use isolated ILogger instance
- No longer modifies global Log.Logger, preventing cross-test
interference
- Tests now use `_logCapture.Logger.Information()` instead of
`Log.Information()`

## Test Plan

- [x] All pre-commit hooks pass (prettier, markdownlint, cspell,
secretlint)
- [x] Build succeeds with no warnings: `dotnet build -c Release
-warnaserror`
- [x] All 93 tests pass: `dotnet test -c Release`
- [x] Code formatting verified: `dotnet format --verify-no-changes`
- [x] npm lint passes: `npm run lint`
- [x] No secrets in code: `npx secretlint .`
- [x] Mutation testing baseline established (14.44%)
- [x] CI passes on all platforms (Windows, Ubuntu, macOS)

## Commits

1. `feat: add release preparation documentation and tooling` - Main docs
and config
2. `fix: use conventional commit prefixes in dependabot config` - CI fix
3. `fix: isolate log capture from global serilog logger` - Test
isolation

Closes #15

🤖 Generated with [Claude Code](https://claude.com/claude-code)

---------

Co-authored-by: Claude (Tech Lead) <[email protected]>
Co-authored-by: Claude Opus 4.5 <[email protected]>

Author: 3 people

.config/dotnet-tools.json

@@ -8,7 +8,7 @@
       "rollForward": false
     },
     "dotnet-stryker": {
-      "version": "4.4.2",
+      "version": "4.11.0",
       "commands": ["dotnet-stryker"],
       "rollForward": false
     },

.cspell.json

@@ -163,7 +163,8 @@
     "innersource",
     "amannn",
     "msbuild",
-    "msdata"
+    "msdata",
+    "unlist"
   ],
   "flagWords": [],
   "allowCompoundWords": true,

.cspell/project-terms.txt

@@ -1,3 +1,6 @@
 # Project-specific terms for spell checking
 # Add words here that are specific to this project
 # One word per line
+
+# NuGet terminology
+unlist

.github/dependabot.yml

@@ -13,7 +13,8 @@ updates:
       timezone: 'UTC'
     open-pull-requests-limit: 10
     commit-message:
-      prefix: 'deps'
+      prefix: 'build'
+      prefix-development: 'build'
       include: 'scope'
     groups:
       # Group all patch updates together
@@ -60,7 +61,8 @@ updates:
       timezone: 'UTC'
     open-pull-requests-limit: 5
     commit-message:
-      prefix: 'deps'
+      prefix: 'build'
+      prefix-development: 'build'
       include: 'scope'
     groups:
       # Group all npm patch updates

CHANGELOG.md

@@ -7,19 +7,36 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
-### Features
-
-- Initial project structure and tooling
-- .NET 10 project configuration
-- GitHub Actions CI/CD pipeline
-- Comprehensive documentation framework
+### Added
+
+- Core application functionality
+  - PTY-based Claude CLI monitoring with `ClaudeMonitor`
+  - Rate limit detection with configurable patterns
+  - Auto-resume after rate limit wait period
+  - Headless mode with auto-response to prompts
+  - Configurable wait times and patterns via `WrapperConfig`
+- CLI interface with argument parsing
+  - `--help`, `--version`, `--diagnose` flags
+  - `--headless` and `--dangerously-skip-permissions` for automation
+  - `--prompt` and `--continue` for session management
+  - `--wait` for custom rate limit wait times
+  - `--verbose` for detailed logging
+- Comprehensive test coverage (93 tests)
+  - Unit tests for all components
+  - BDD feature tests (rate limit detection, auto-resume, headless mode)
+  - E2E tests for application startup
+  - Architecture tests for dependency validation
+- .NET 10 project structure with central package management
+- GitHub Actions CI/CD pipeline (build, test, lint on Windows/macOS/Linux)
+- Comprehensive documentation framework with Docusaurus
 - Pre-commit hooks and code quality enforcement
+- 32 Architecture Decision Records (ADRs)
 
 ### Documentation
 
-- Architecture Decision Records (ADRs)
+- Developer onboarding guides
 - Contributing guidelines
-- Developer playbooks
 - Accessibility standards
+- Release playbooks
 
 <!-- generated by git-cliff -->

README.md

@@ -1,5 +1,9 @@
 # Claude Auto Resume
 
+[![CI](https://github.com/mcj-coder-org/claude-auto-resume/actions/workflows/ci.yml/badge.svg)](https://github.com/mcj-coder-org/claude-auto-resume/actions/workflows/ci.yml)
+[![.NET 10](https://img.shields.io/badge/.NET-10-512BD4)](https://dotnet.microsoft.com/)
+[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)
+
 A cross-platform .NET tool that wraps the Claude CLI to provide automatic session
 resumption when rate limits are hit.
 
@@ -14,67 +18,87 @@ limit resets.
 
 - **Automatic rate limit detection** - Monitors Claude CLI output for rate limit messages
 - **Seamless resumption** - Waits for the reset time and automatically resumes your session
+- **Headless mode** - Run unattended with auto-response to prompts
 - **Transparent operation** - Acts as a passthrough until rate limits are detected
 - **Cross-platform** - Runs on Windows, macOS, and Linux
 - **Privacy-first** - No telemetry or data collection (see [Privacy Policy](docs/standards/privacy.md))
 
 ## Installation
 
+### Prerequisites
+
+- [Claude CLI](https://docs.anthropic.com/en/docs/agents-and-tools/claude-code/overview) installed and in PATH
+- .NET 10 Runtime (for .NET tool installation)
+
 ### .NET Tool (Recommended)
 
-````bash
+```bash
 dotnet tool install -g McjCoderOrg.ClaudeAutoResume
-```text
+```
 
 ### Standalone Executable
 
-Download the latest release for your platform from the [Releases](https://github.com/mcj-coder-org/claude-auto-resume/releases) page.
-
-## Usage
+Download the latest release for your platform from the
+[Releases](https://github.com/mcj-coder-org/claude-auto-resume/releases) page.
 
-Use `claude-auto-resume` as a drop-in replacement for `claude`:
+## Quick Start
 
 ```bash
-# Basic usage
+# Interactive mode (default) - use like normal claude
 claude-auto-resume
 
-# With verbose logging
-claude-auto-resume --verbose
+# With an initial prompt
+claude-auto-resume -p "implement the login feature"
 
-# Run diagnostics
-claude-auto-resume --diagnose
-
-# Pass arguments to Claude CLI
-claude-auto-resume -- --help
-```text
+# Continue previous session
+claude-auto-resume -c -p "continue where we left off"
 
-## Configuration
+# Headless mode for CI/automation
+claude-auto-resume --headless --dangerous -p "implement feature"
 
-Configuration can be provided via CLI arguments, environment variables, project config, or user config.
+# Pass additional args to claude
+claude-auto-resume -- --model claude-3-opus
+```
 
-### CLI Arguments
+## Usage
 
-```bash
-claude-auto-resume --retry-delay 10000 --max-retries 5
 ```text
+USAGE:
+    claude-auto-resume [OPTIONS] [-- CLAUDE_ARGS...]
 
-### Environment Variables
+OPTIONS:
+    -h, --help                      Show help
+    -v, --version                   Show version information
+    -p, --prompt <PROMPT>           Initial prompt to send to Claude
+    -c, --continue                  Continue previous conversation
+    -w, --wait <MINUTES>            Minutes to wait on rate limit (default: 15)
+    -V, --verbose                   Enable verbose logging to file
+    --headless                      Run without user input (auto-respond to prompts)
+    --dangerously-skip-permissions  Pass dangerous flag to Claude (required for headless)
+    --dangerous                     Alias for --dangerously-skip-permissions
+    --diagnose                      Run environment diagnostics
 
-```bash
-export CLAUDE_AUTO_RESUME_RETRY_DELAY=10000
-export CLAUDE_AUTO_RESUME_MAX_RETRIES=5
-```text
+ENVIRONMENT:
+    CLAUDE_WAIT_MINUTES             Override default wait time
+```
 
-### Configuration File
+## Modes
 
-Create `.claude-auto-resume.json` in your project directory or `~/.config/claude-auto-resume/config.json` for user-level config:
+### Interactive (Default)
 
-```json
-{
-  "retryDelayMs": 10000,
-  "maxRetries": 5
-}
-```text
+- Full PTY pass-through with colors
+- You type, Claude responds
+- Auto-waits and continues on rate limit
+
+### Headless (`--headless --dangerous`)
+
+- No user input required
+- Auto-responds 'y' to permission prompts
+- Detects when Claude hangs waiting for input
+- Ideal for CI/CD pipelines
+
+> **Warning:** `--dangerously-skip-permissions` allows Claude to execute commands
+> without confirmation. Use only in trusted environments.
 
 ## Development
 
@@ -99,11 +123,12 @@ dotnet build
 
 # Run tests
 dotnet test
-```text
+```
 
 ### For AI Agents
 
-If you're an AI agent working on this codebase, start by reading [AGENTS.md](AGENTS.md) for orientation and documentation routing guidance.
+If you're an AI agent working on this codebase, start by reading [AGENTS.md](AGENTS.md)
+for orientation and documentation routing guidance.
 
 ### Documentation
 
@@ -130,30 +155,33 @@ If you're an AI agent working on this codebase, start by reading [AGENTS.md](AGE
                          │
 ┌────────────────────────▼────────────────────────────┐
 │            claude-auto-resume (this tool)           │
-│  • Wraps Claude CLI process                         │
+│  • Wraps Claude CLI in pseudo-terminal              │
 │  • Monitors output for rate limits                  │
 │  • Auto-resumes when limit resets                   │
+│  • Handles prompts in headless mode                 │
 └────────────────────────┬────────────────────────────┘
                          │
 ┌────────────────────────▼────────────────────────────┐
 │               Claude CLI (Anthropic)                 │
 └─────────────────────────────────────────────────────┘
-```text
+```
 
 ## Exit Codes
 
-| Code | Meaning                                      |
-| ---- | -------------------------------------------- |
-| 0    | Success                                      |
-| 1    | General error                                |
-| 2    | Configuration error                          |
-| 3    | Claude CLI not found                         |
-| 4    | Rate limit detected (when not auto-resuming) |
-| 5    | User cancelled (Ctrl+C)                      |
+| Code | Meaning                 |
+| ---- | ----------------------- |
+| 0    | Success                 |
+| 1    | General error           |
+| 2    | Invalid arguments       |
+| 3    | Configuration error     |
+| 4    | Claude CLI not found    |
+| 5    | Rate limit detected     |
+| 6    | User cancelled (Ctrl+C) |
 
 ## Privacy
 
-This tool collects **no telemetry, analytics, or usage data**. All data remains on your machine. See our [Privacy Policy](docs/standards/privacy.md) for details.
+This tool collects **no telemetry, analytics, or usage data**. All data remains on your
+machine. See our [Privacy Policy](docs/standards/privacy.md) for details.
 
 ## License
 
@@ -163,4 +191,3 @@ This tool collects **no telemetry, analytics, or usage data**. All data remains
 
 - [Anthropic](https://www.anthropic.com/) for the Claude CLI
 - All contributors to this project
-````

docs/adr/0031-pr-validation-automation.md

@@ -4,7 +4,7 @@ description: |
   Automated PR validation using DangerJS to enforce Definition of Done
   checklists, evidence requirements, and design plan archiving.
 decision: Use DangerJS for automated PR validation with custom rules
-status: proposed
+status: accepted
 audience: [developer, agent]
 topics: [ci, automation, pr, validation, dangerjs]
 last_validated: 2026-01-11
@@ -14,7 +14,7 @@ last_validated: 2026-01-11
 
 ## Status
 
-Proposed
+Accepted
 
 ## Context