setup

Compound Engineering Setup

Interactive setup for compound-engineering.local.md — configures which agents run during /workflows:review and /workflows:work.

Step 1: Check Existing Config

Read compound-engineering.local.md in the project root. If it exists, display current settings summary and use AskUserQuestion:

question: "Settings file already exists. What would you like to do?"
header: "Config"
options:
  - label: "Reconfigure"
    description: "Run the interactive setup again from scratch"
  - label: "View current"
    description: "Show the file contents, then stop"
  - label: "Cancel"
    description: "Keep current settings"

If "View current": read and display the file, then stop. If "Cancel": stop.

Step 2: Detect Python Stack

Auto-detect the project stack by scanning dependency files:

Detection sources: pyproject.toml, requirements.txt, setup.py, setup.cfg, Pipfile

Framework detection:

Dependency	Detected As
django	Django
djangorestframework	DRF
django-ninja	Django Ninja
celery	Celery
django-channels or channels	Channels
fastapi	FastAPI
sqlalchemy	SQLAlchemy
psycopg2, psycopg, asyncpg	PostgreSQL
django.contrib.gis in settings or geoalchemy2	PostGIS
pydantic	Pydantic
alembic	Alembic

Confirmation signals:

File/Directory	Confirms
manage.py	Django project
alembic.ini or alembic/	SQLAlchemy + Alembic
celeryconfig.py or celery.py in project	Celery

Use AskUserQuestion:

question: "Detected Python project with: {detected_frameworks}. How would you like to configure?"
header: "Setup"
options:
  - label: "Auto-configure (Recommended)"
    description: "Use smart defaults for your detected stack. Done in one click."
  - label: "Customize"
    description: "Choose frameworks, focus areas, and review depth."

If Auto-configure → Skip to Step 4 with defaults

Always included: [python-style-reviewer, python-typing-reviewer, pytest-reviewer, python-lint, code-simplicity-reviewer, security-sentinel, performance-oracle]

Conditional additions:

Detected	Agents Added
Django	django-reviewer, django-migration-reviewer
DRF	django-drf-reviewer
Django Ninja	(covered by django-reviewer)
Celery	django-celery-reviewer
FastAPI	fastapi-reviewer
SQLAlchemy	fastapi-sqlalchemy-reviewer
PostgreSQL	postgresql-reviewer
PostGIS	postgis-reviewer

If Customize → Step 3

Step 3: Customize (4 questions)

a. Frameworks — confirm or override:

question: "Which frameworks does this project use?"
header: "Frameworks"
multiSelect: true
options:
  - label: "Django"
    description: "Django web framework — adds Django pattern and migration reviewers"
  - label: "DRF"
    description: "Django REST Framework — adds serializer and viewset reviewer"
  - label: "Celery"
    description: "Celery task queue — adds task design and reliability reviewer"
  - label: "FastAPI"
    description: "FastAPI — adds route, DI, and async correctness reviewer"

b. Database — select database tools:

question: "Which database tools apply?"
header: "Database"
multiSelect: true
options:
  - label: "PostgreSQL"
    description: "PostgreSQL — adds query optimization and schema reviewer"
  - label: "PostGIS"
    description: "PostGIS spatial — adds SRID, spatial index, and GeoDjango reviewer"
  - label: "SQLAlchemy"
    description: "SQLAlchemy ORM — adds session lifecycle and Alembic migration reviewer"

c. Focus areas — multiSelect:

question: "Which review areas matter most?"
header: "Focus"
multiSelect: true
options:
  - label: "Security"
    description: "Vulnerability scanning, auth, input validation (security-sentinel)"
  - label: "Performance"
    description: "N+1 queries, memory leaks, complexity (performance-oracle)"
  - label: "Architecture"
    description: "Design patterns, SOLID, separation of concerns (architecture-strategist)"
  - label: "Code simplicity"
    description: "Over-engineering, YAGNI violations (code-simplicity-reviewer)"

d. Depth:

question: "How thorough should reviews be?"
header: "Depth"
options:
  - label: "Thorough (Recommended)"
    description: "Stack reviewers + all selected focus agents."
  - label: "Fast"
    description: "Stack reviewers + code simplicity only. Less context, quicker."
  - label: "Comprehensive"
    description: "All above + git history, data integrity, agent-native checks."

e. Workflow Skills — configure per-workflow skills:

question: "Load framework/database skills into workflow commands?"
header: "Wf Skills"
options:
  - label: "Auto-configure (Recommended)"
    description: "Load detected stack skills into all workflows automatically."
  - label: "Customize per workflow"
    description: "Choose which skills load for each workflow command."
  - label: "Skip"
    description: "Don't configure workflow skills. Commands use only their built-in skills."

If Auto-configure: Build a workflow_skills map using the detected/selected frameworks and database:

Detected	Skills Added
Django	django-patterns
DRF	django-drf
Celery	django-async
FastAPI	fastapi-patterns
SQLAlchemy	fastapi-sqlalchemy
PostgreSQL	postgresql-patterns
PostGIS	postgis-patterns

Apply these skills to all 5 workflow commands uniformly. Always add pytest-patterns to work and review.

If Customize per workflow: For each of the 5 workflows (brainstorm, plan, work, review, compound), present a multiSelect of available skills:

question: "Which skills should load for /workflows:{command}?"
header: "{Command} Skills"
multiSelect: true
options:
  - label: "django-patterns"
    description: "Django models, views, admin, signals, middleware patterns"
  - label: "django-drf"
    description: "DRF serializers, viewsets, permissions"
  - label: "django-async"
    description: "Celery task patterns, Django Channels"
  - label: "fastapi-patterns"
    description: "FastAPI routes, DI, async patterns"
  - label: "fastapi-sqlalchemy"
    description: "SQLAlchemy 2.0 async, Alembic migrations"
  - label: "pytest-patterns"
    description: "Fixtures, parametrize, mocking, async testing"
  - label: "postgresql-patterns"
    description: "Schema design, indexing, query optimization"
  - label: "postgis-patterns"
    description: "Spatial queries, GeoDjango, GeoAlchemy2"

Also include any project-local skills found in .claude/skills/ as additional options.

If Skip: Omit workflow_skills from the config file entirely.

Step 4: Build Agent List and Write File

Base agents (always included): python-style-reviewer, python-typing-reviewer, pytest-reviewer, python-lint

Framework agents:

• Django → django-reviewer, django-migration-reviewer
• DRF → django-drf-reviewer
• Celery → django-celery-reviewer
• FastAPI → fastapi-reviewer
• SQLAlchemy → fastapi-sqlalchemy-reviewer

Database agents:

• PostgreSQL → postgresql-reviewer
• PostGIS → postgis-reviewer

Focus area agents:

• Security → security-sentinel
• Performance → performance-oracle
• Architecture → architecture-strategist
• Code simplicity → code-simplicity-reviewer

Depth:

• Thorough: stack + selected focus areas
• Fast: stack + code-simplicity-reviewer only
• Comprehensive: all above + git-history-analyzer, data-integrity-guardian, agent-native-reviewer

Plan review agents: base Python reviewers + code-simplicity-reviewer.

Write compound-engineering.local.md:

---
review_agents: [{computed agent list}]
plan_review_agents: [{computed plan agent list}]
frameworks: [{detected or selected frameworks}]
database: [{detected or selected database tools}]
workflow_skills:
  brainstorm: [{computed skill list}]
  plan: [{computed skill list}]
  work: [{computed skill list}]
  review: [{computed skill list}]
  compound: [{computed skill list}]
---

# Review Context

Add project-specific review instructions here.
These notes are passed to all review agents during /workflows:review and /workflows:work.

Examples:
- "We use Django Ninja for the API — check schema design and router organization"
- "Our API is public — extra scrutiny on input validation"
- "Performance-critical: we serve 10k req/s on this endpoint"
- "We use Celery with Redis broker — check task idempotency"

Omit the workflow_skills block entirely if the user chose "Skip" in Step 3e.

Step 5: Confirm

Saved to compound-engineering.local.md

Stack:        Python
Frameworks:   {framework list}
Database:     {database list}
Review depth: {depth}
Agents:       {count} configured
              {agent list, one per line}
Workflow skills: {count} skills across {count} workflows
              brainstorm: {skill list or "none"}
              plan:       {skill list or "none"}
              work:       {skill list or "none"}
              review:     {skill list or "none"}
              compound:   {skill list or "none"}

Tip: Edit the "Review Context" section to add project-specific instructions.
     Edit workflow_skills in the YAML frontmatter to customize per-workflow skills.
     Re-run this setup anytime to reconfigure.