feat: implement docker container manager mvp (#25)

Author: remcostoeten

.agent/AGENTS.md

@@ -1,16 +1,18 @@
 # AI Agent Guidelines
 
 ## Release Notes & Changelogs
+
 - **Location**: Release notes drafts must be written to `docs/RELEASE_NOTES.md`.
 - **Format**: Plain text Markdown. **STRICTLY NO EMOJIS**. No "marketing fluff".
 - **Structure**:
-  - Highlights (Critical changes only)
-  - Features (New capabilities)
-  - Fixes (Bug fixes)
-  - Technical (Refactors, build changes)
+    - Highlights (Critical changes only)
+    - Features (New capabilities)
+    - Fixes (Bug fixes)
+    - Technical (Refactors, build changes)
 - **Version**: Do not invent version numbers. Use the provided one or increment logical patch/minor.
 
 ## Documentation
+
 - Keep `README.md` professional.
 - **NO DECORATIVE EMOJIS**.
 - Use text labels (Done/WIP) instead of status icons.

.agent/claude.md

@@ -1,18 +1,22 @@
 # Claude Context
 
 ## Project Structure
+
 - **Root**: `/home/remco-stoeten/dora`
 - **CLI**: `tools/dora-cli` (Go)
 - **Scripts**: `tools/scripts` (TypeScript)
 - **Docs**: `docs/` (Release notes, changelogs)
 
 ## Key Rules
+
 1. **No Emojis**: Never use emojis in technical documentation or release notes.
 2. **Path Awareness**: Always check `docs/` for release notes, not root.
 3. **CLI**: The `dora` binary is the source of truth for build management.
 
 ## Release Process
+
 When asked to handle a release:
+
 1. Check `docs/RELEASE_NOTES.md` for the current draft.
 2. Use `dora-cli` scripts if available (`bun run generate-release`).
 3. If writing notes manually, follow `AGENTS.md` guidelines.

.agent/rules/no-arrow-constants.md

@@ -7,25 +7,30 @@ RULE: Function Declarations Only
 The codebase must not contain arrow functions.
 
 Disallow:
+
 - const fn = () => {}
 - export const fn = () => {}
 - use of => in any form
 
 Allow only:
+
 - function fn() {}
 - export function fn() {}
 
 Enforcement:
+
 - If any arrow function is detected, reject the output.
 - Rewrite all arrow functions into standard function declarations.
 - Do not introduce arrow functions in new code.
 - Do not suggest arrow functions in examples or explanations.
 
 Rationale:
+
 - Enforce consistent style.
 - Improve stack traces and debugging.
 - Avoid lexical this and implicit returns.
 
 Compliance check:
+
 - Scan for "=>" before returning any code.
 - If found, block the response and regenerate using function declarations only.

.agent/rules/type-convention.md

@@ -4,35 +4,34 @@ trigger: always_on
 
 Component Props Typing — Mandatory `Props` for Single Type Files
 
-
 When defining a React/Next.js component with props, if the file contains **only one non-exported type**, it **MUST** be named exactly `TProps`.
- 
+
 Example:
 For this component:
- 
+
 """tsx
 function Button({ name, onClick, another }) {
-  // ...
+// ...
 }
 """
- 
+
 You MUST define props like this:
 """tsx
 type Props = {
-  name: string;
-  onClick?: () => void;
-  another?: any;
+name: string;
+onClick?: () => void;
+another?: any;
 };
- 
+
 export function Button({ name, onClick, another }: Props) {
-  // ...
+// ...
 }
 """
 Key points:
- 
+
 - TProps is always used if there’s only one type in the file.
 - TProps must not be exported — it is local to the component file.
 - Props must be explicitly typed — never leave props untyped or inline-typed.
 - Follow camelCase naming for properties inside TProps, but the type name itself must be TProps.
- 
-This ensures consistent, clear, and easy-to-find component prop typings across the codebase.
+
+This ensures consistent, clear, and easy-to-find component prop typings across the codebase.

.agent/workflows/release.md

@@ -5,6 +5,7 @@ description: Generate release notes and changelog using AI
 # Release Generation Workflow
 
 ## Prerequisites
+
 1. Ensure you have a `.env` file with your API key configured (see `tools/scripts/.env.example`)
 2. For Gemini: Set `GEMINI_API_KEY` in your environment
 3. For Ollama: Run `bun setup:ai` first to verify your local setup
@@ -14,43 +15,51 @@ description: Generate release notes and changelog using AI
 // turbo-all
 
 ### Test API Connection
+
 ```bash
 bun release:gen --test
 ```
 
 ### List Available Models (Gemini only)
+
 ```bash
 bun release:gen --list-models
 ```
 
 ### Preview Release Notes (Dry Run)
+
 ```bash
 bun release:gen --dry-run
 ```
 
 ### Generate Release Notes
+
 ```bash
 bun release:gen
 ```
 
 ### Generate and Build Executables
+
 ```bash
 bun release:gen --build
 ```
 
 ### Generate with Version Bump
+
 ```bash
 bun release:gen --version-bump=patch
 bun release:gen --version-bump=minor
 bun release:gen --version-bump=major
 ```
 
 ## Output Files
+
 - `RELEASE_NOTES.md` - Full release notes
 - `CHANGELOG.md` - Appended changelog entry
 - `apps/desktop/package.json` - Updated version (if using --version-bump)
 
 ## Troubleshooting
+
 1. **API Error**: Run `--test` to verify connectivity
 2. **Bad Output**: The AI should return valid JSON; if not, check model availability with `--list-models`
 3. **No Commits**: Ensure you have commits since the last git tag

.agent/workflows/setup-ai.md

@@ -5,6 +5,7 @@ description: Setup and verify local Ollama AI
 # Local AI Setup Workflow
 
 ## Prerequisites
+
 1. Install Ollama from https://ollama.com/download
 2. Start the Ollama server: `ollama serve`
 
@@ -13,37 +14,42 @@ description: Setup and verify local Ollama AI
 // turbo-all
 
 ### Basic Setup
+
 ```bash
 bun setup:ai
 ```
 
 ### Specify a Different Model
+
 ```bash
 bun setup:ai --model mistral
 bun setup:ai --model codellama
 bun setup:ai --model llama3:8b
 ```
 
 ### Diagnose Connection Issues
+
 ```bash
 bun setup:ai --diagnose
 ```
 
 ### Show Help
+
 ```bash
 bun setup:ai --help
 ```
 
 ## Popular Models for Release Notes
 
-| Model | Size | Speed | Quality |
-|-------|------|-------|---------|
-| `llama3` | 4.7GB | Fast | Good |
-| `llama3:70b` | 40GB | Slow | Excellent |
-| `mistral` | 4.1GB | Fast | Good |
-| `codellama` | 3.8GB | Fast | Code-focused |
+| Model        | Size  | Speed | Quality      |
+| ------------ | ----- | ----- | ------------ |
+| `llama3`     | 4.7GB | Fast  | Good         |
+| `llama3:70b` | 40GB  | Slow  | Excellent    |
+| `mistral`    | 4.1GB | Fast  | Good         |
+| `codellama`  | 3.8GB | Fast  | Code-focused |
 
 ## Troubleshooting
+
 1. **Connection refused**: Make sure `ollama serve` is running
 2. **Model not found**: Run `ollama pull <model-name>`
 3. **Remote server**: Set `OLLAMA_HOST` and `OLLAMA_PORT` env vars