dotPilot: Skill for Codex

MCAF: ADR Writing

Trigger On

• a dependency, boundary, platform, contract, or data model is changing
• a design decision has meaningful trade-offs that should be recorded
• a repo-wide engineering policy needs a durable rationale

Value

• produce a concrete project delta: code, docs, config, tests, CI, or review artifact
• reduce ambiguity through explicit planning, verification, and final validation skills
• leave reusable project context so future tasks are faster and safer

Do Not Use For

• feature-level behaviour details without an architecture decision
• generic architecture overview content

Inputs

• docs/Architecture.md
• related feature docs
• the nearest AGENTS.md
• current constraints, options, and risks

Quick Start

1. Read the nearest AGENTS.md and confirm scope and constraints.
2. Run this skill's Workflow through the Ralph Loop until outcomes are acceptable.
3. Return the Required Result Format with concrete artifacts and verification evidence.

Workflow

1. Start from the concrete decision that must be made now.
2. If the ADR is missing, scaffold it from references/adr-template.md.
3. Record:
   • context and problem
   • chosen decision
   • alternatives considered
   • trade-offs and consequences
   • implementation plan
4. Add diagrams only when they remove ambiguity.
5. Link the ADR to affected feature docs and docs/Architecture.md.

Deliver

• docs/ADR/ADR-XXXX-short-title.md
• linked updates to architecture docs when the decision changes boundaries

Validate

• the decision and rejected alternatives are explicit
• trade-offs are concrete, not hand-wavy
• implementation impact is clear
• a future engineer can understand why this path was chosen

Ralph Loop

Use the Ralph Loop for every task, including docs, architecture, testing, and tooling work.

1. Brainstorm first (mandatory):
   • analyze current state
   • define the problem, target outcome, constraints, and risks
   • generate options and think through trade-offs before committing
   • capture the recommended direction and open questions
2. Plan second (mandatory):
   • write a detailed execution plan from the chosen direction
   • list final validation skills to run at the end, with order and reason
3. Execute one planned step and produce a concrete delta.
4. Review the result and capture findings with actionable next fixes.
5. Apply fixes in small batches and rerun the relevant checks or review steps.
6. Update the plan after each iteration.
7. Repeat until outcomes are acceptable or only explicit exceptions remain.
8. If a dependency is missing, bootstrap it or return status: not_applicable with explicit reason and fallback path.