CDD Pipeline

This project calls its release workflow Context Driven Development (CDD)

Workspace Configuration

All CDD commands and scripts require the

$CDD_DIR

environment variable.

export CDD_DIR=/path/to/workspace

This allows parallel workspaces without VCS conflicts. Each workspace is independent - different branches can use different workspace directories simultaneously.

Scripts will abort with a clear error if

$CDD_DIR

is not set.

Core Philosophy

Use user/agent interactive development time during the Research and Plan phases to prepare tightly focused work that pre-loads the exact required context to run autonomously during the Execute phase.

Pipeline Phases

• Research
• Plan
• Execute

1. Research

What: Build foundational understanding before design decisions.

Artifacts:

• $CDD_DIR/README.md

  - High-level description of release features

• $CDD_DIR/user-stories/

  - User actions and expected system responses

• $CDD_DIR/research/

  - Facts gathered from internet, docs, codebase

Why this phase exists: Decisions made without research are assumptions. Research surfaces constraints, prior art, and edge cases that would otherwise appear during execution (when fixing is expensive).

Key insight: Research artifacts become reference material for execution. Sub-agents can read

$CDD_DIR/research/

without re-fetching from the internet. The research phase pays the lookup cost once.

2. Plan

What: Transform research into implementation decisions.

Artifacts:

• $CDD_DIR/plan/

  - Implementation decisions, architecture choices, detailed designs

Why this phase exists: Planning is the translation layer between "what we want" (research) and "what to do". It makes architectural decisions explicit.

3. Verify (Sub-phase of Plan)

What: Iterative verification of plan alignment.

Artifacts:

• $CDD_DIR/gap.md

  - List of issues to investigate and resolve

• $CDD_DIR/verified.md

  - Log of issues that have been fixed

Verification:

/cdd:gap  # Verify README/user-stories → plan alignment, writes gap.md
fix gaps  # Address issues in gap.md, update plan documents
update verified.md  # Log what was fixed with references to gap.md
/cdd:gap  # Repeat until clean

Checks alignment down the pyramid:

• README/research/user-stories → plan
• Naming, error handling, memory management conventions
• Integration points, build changes, migrations

Why this phase exists: This is the last moment of full context visibility. Cross-cutting issues and inconsistencies can only be caught while the complete picture is visible.

Key insights:

• Specific checklists prevent vague "find issues" prompts

• gap.md

  contains current issues to investigate

• verified.md

  tracks what's been fixed, prevents re-checking completed items

Convergence pattern: Each gap-finding run identifies issues and writes gap.md. After fixing, log in verified.md and re-run. When no substantive gaps remain, verification is complete.

4. Execute

What: Unattended execution via the Ralph harness script.

Artifacts:

• $CDD_DIR/goals/*-goal.md

  - Goal file(s) for ralph loop execution

• $CDD_DIR/goals/*-progress.jsonl

  - Auto-generated progress log

• $CDD_DIR/goals/*-summary.md

  - Auto-generated progress summary

Ralph script:

.claude/harness/ralph/run \
  --goal=$CDD_DIR/goals/implementation-goal.md \
  --duration=4h \
  --model=sonnet \
  --reasoning=low

Process: Ralph iteratively works toward the goal until DONE or time expires:

1. Read goal + progress history
2. Make incremental progress (code changes, tests, fixes)
3. Commit progress via

   jj commit

4. Append progress to JSONL file
5. Repeat until objective achieved

Why Ralph: The plan phase prepared complete context. Ralph execution is mechanical - read plan, implement, test, commit. No research, no exploration, just steady progress toward the goal using the plan as a reference.

Post-execution: After Ralph completes (or during iterations), quality checks and refactoring may be needed:

ralph → /check:quality → /refactor → /check:quality

Directory Structure

$CDD_DIR/
├── README.md          # High-level release description (PRD)
├── user-stories/      # User actions and system responses
│   └── README.md      # Index/overview
├── research/          # Internet facts for reference
│   └── README.md      # Index/overview
├── plan/              # Implementation decisions
│   └── README.md      # Index/overview
├── gap.md             # Current issues to investigate and resolve
├── verified.md        # Log of issues that have been fixed
├── goals/             # Ralph goal files and execution state
│   ├── *-goal.md      # Goal definition(s)
│   ├── *-progress.jsonl   # Progress tracking (auto-generated)
│   └── *-summary.md   # Progress summary (auto-generated)
├── tmp/               # Temp files during execution
└── ...                # Other files permitted, ignored by pipeline

Abstraction Levels

Each directory serves a distinct abstraction level. Content should not leak between levels.

$CDD_DIR/plan/ - Coordination Layer

Plan documents coordinate everything shared.

The plan owns:

• Public symbols - Function names, struct names, enum names
• Function signatures - Arguments, argument order, return types
• Struct members - Field names, field order, field types
• Enums - Value names and meanings
• Allowed libraries - Which dependencies can be used, which are forbidden
• Architectural choices - Module boundaries, data flow, ownership rules

The plan is the contract.

DO NOT include: Implementation code, detailed algorithms, function bodies.

Artifact Authority

Artifacts form a hierarchy of authority:

$CDD_DIR/README.md (authoritative)
        ↓
user-stories + research (derived)
        ↓
      plan (derived)

Rules:

1. Lower-level artifacts derive from higher-level ones
2. Contradictions between levels must be resolved immediately
3. Resolution is always top-down: align lower artifacts to higher authority
4. Higher-level documents may change through iteration—but when they do, all derived artifacts must be reviewed and realigned

Why this matters: Consistency is enforced during authoring because it cannot be enforced during execution.

Writing Ralph Goal Files

After creating plan documents, write goal files in

$CDD_DIR/goals/*-goal.md

.

Use

/load goal-authoring

for detailed guidance. Key points:

• Ralph has unlimited context through iteration - reference all relevant docs
• Specify WHAT to achieve (outcomes), never HOW (steps/order)
• One cohesive objective per goal - trust Ralph to discover the path
• Complete acceptance criteria - Ralph needs to know when done

Lifecycle

Create empty $CDD_DIR/ → Research → Plan → Verify → Execute → Delete $CDD_DIR/

The workspace directory is ephemeral. It exists only for the duration of the release workflow. After successful execution, it's deleted. The work lives in the codebase; the pipeline artifacts are disposable.

Efficiency Principles

1. Load skills on-demand - Don't preload "just in case"
2. Complete authoring - Spend tokens researching during authoring so execution doesn't need to
3. Reference vs working knowledge - Large docs go in separate skills, loaded when needed

Design Tradeoffs

Decision	Tradeoff	Rationale
Full context in verify	Expensive per-run	Last chance to catch cross-cutting issues
Verification is human-terminated	Subjective	Models don't converge to "done" naturally
Research cached in files	Stale if release spans days	Avoids re-fetching during execution

When to Load This Skill

• Designing or modifying the release workflow
• Debugging why execution sub-agents are failing
• Optimizing token usage in the pipeline
• Understanding why a phase exists before changing it