# Agentic Engineering Toolkit

A minimal, portable toolkit of practices, templates, and skills for rigorous AI-assisted development. Built for personal use across any project and any CLI-based agent harness (Claude Code, pi, opencode, etc.).

The goal: have a single, complete setup that can be dropped into any project to enforce disciplined agentic engineering — TDD, adversarial review, fresh contexts, and human-in-the-loop commits — without reinventing the wheel each time.

## Workflow

1. **Init** — Set up a repo with `context-init` (project documentation) and `agent-container` (isolated dev environment)
2. **Work** — Launch the agent container with a task, follow the core loop: spec, failing tests, minimal implementation, adversarial review, commit
3. **Maintain** — Review state of work, clean up, update documentation

## Structure

- `skills/` — Portable markdown instruction sets. Each is a self-contained practice (with a `SKILL.md`) that can be loaded into any agent harness.
- `bin/` — Standalone scripts (e.g., adversarial review via a second model's API).

## The Core Loop

```
  Spec ──→ Failing Tests ──→ Minimal Implementation ──→ All Tests Pass
   ↑                                                          │
   │                                                          ↓
   │                                              Adversarial Review
   │                                              (fresh context, hostile)
   │                                                          │
   │         ┌─── hallucinating? ──→ DONE (converged) ───→ Commit
   │         │
   └── real findings ──→ Feed back to Builder (fresh session)
```

**Exit condition:** When the adversary starts inventing problems that don't exist in the code, you've converged. Ship it.

## Principles

1. **Specs are truth.** Tests serve specs. Code serves tests.
2. **Tests before code.** No implementation without a failing test.
3. **Fresh context per task.** Long sessions degrade. Commit and restart.
4. **Different model for review.** Cognitive diversity catches blind spots.
5. **Fresh context per review.** No relationship drift. Every roast is the first roast.
6. **Hallucination = exit signal.** When the adversary makes things up, you're done.
7. **You commit manually.** The agent proposes. You dispose.