Paths: File paths (shared/ , references/ , ../ln-* ) are relative to skills repo root. If not found at CWD, locate this SKILL.md directory and go up one level for repo root. If shared/ is missing, fetch files via WebFetch from https://raw.githubusercontent.com/levnikolaevich/claude-code-skills/master/skills/{path} .

DevOps Documentation Creator

Type: L3 Worker

L3 Worker that creates infrastructure.md and runbook.md. Infrastructure inventory is always created; runbook is conditional on Docker presence.

Purpose & Scope

• Creates infrastructure.md (always) — declarative inventory: WHAT is deployed WHERE

• Creates runbook.md (if hasDocker) — procedural guide: HOW to deploy/restart/troubleshoot

• Receives Context Store from ln-110-project-docs-coordinator

• Never gathers context itself; uses coordinator input

Invocation (who/when)

• ln-110-project-docs-coordinator: ALWAYS invoked (infrastructure.md is unconditional)

• Never called directly by users

Inputs

From coordinator:

• contextStore : Context Store with DevOps-specific data

• DOCKER_COMPOSE_DEV (development setup)

• DOCKER_COMPOSE_PROD (production setup)

• ENV_VARIABLES (from .env.example)

• STARTUP_SEQUENCE (services order)

• DEPLOYMENT_TARGET (AWS, Vercel, Heroku)

• CI_CD_PIPELINE (from .github/workflows)

• DOCKER_SERVICES (parsed from docker-compose.yml services)

• DEPLOYMENT_SCALE ("single" | "multi" | "auto-scaling" | "gpu-based")

• DEVOPS_CONTACTS (from CODEOWNERS, package.json author, git log)

• HAS_GPU (detected from docker-compose nvidia runtime)

• SERVER_INVENTORY (from SSH config, deploy targets)

• DOMAIN_DNS (from docker-compose VIRTUAL_HOST vars, nginx configs)

• ARTIFACT_REPOSITORY (from .env registry URLs, .npmrc, pip.conf)

• HOST_REQUIREMENTS (from docker-compose deploy.resources.limits)

• targetDir : Project root directory

• flags : { hasDocker }

MANDATORY READ: Load shared/references/docs_quality_contract.md , shared/references/docs_quality_rules.json , and shared/references/markdown_read_protocol.md .

Documents Created (2: 1 always + 1 conditional)

File Condition Questions Auto-Discovery

docs/project/infrastructure.md Always Q52-Q55 Medium

docs/project/runbook.md hasDocker Q46-Q51 High

Workflow

Phase 1: Check Conditions

• Parse flags from coordinator

• infrastructure.md: ALWAYS proceeds (no condition check)

• runbook.md: Create ONLY if hasDocker=true

• If target file already exists: skip that file (idempotent)

Phase 2a: Create infrastructure.md (unconditional)

• Check if docs/project/infrastructure.md exists

• If exists: skip with log

• If not exists:

• Copy references/templates/infrastructure_template.md

• Replace placeholders with Context Store values

• Preserve the shared opening contract and standard top sections from the template

• Populate Server Inventory from SERVER_INVENTORY

• Populate Port Allocation from DOCKER_SERVICES port mappings

• Populate Deployed Services from DOCKER_SERVICES

• Populate CI/CD Pipeline from CI_CD_PIPELINE

• Never leave template markers in published infrastructure docs

• If data is missing: omit the claim or use a concise neutral fallback, but do NOT emit [TBD: ...]

• Conditional Section Pruning:

• If no CI/CD detected: replace with concise empty-state note, not [TBD: ...]

• If no ARTIFACT_REPOSITORY: replace with concise empty-state note, not [TBD: ...]

• If single server / no SERVER_INVENTORY: simplify to single-column table

• If !HAS_GPU: remove GPU column from Server Inventory and Deployed Services

• Populate Deployed Services ONLY from DOCKER_SERVICES (no generic examples)

Phase 2b: Create runbook.md (conditional)

• If !hasDocker : skip entirely

• Check if docs/project/runbook.md exists

• If exists: skip with log

• If not exists:

• Copy references/templates/runbook_template.md

• Replace placeholders with Context Store values

• Preserve the shared opening contract and standard top sections from the template

• Populate setup steps from package.json scripts

• Extract env vars from .env.example

• Never leave template markers in published runbooks

• If data is missing: omit the claim or use a concise neutral fallback, but do NOT emit [TBD: ...]

• Conditional Section Pruning:

• If DEPLOYMENT_SCALE != "multi" or "auto-scaling": Remove scaling/load balancer sections

• If !HAS_GPU: Remove GPU-related sections (nvidia runtime, CUDA)

• If service not in DOCKER_SERVICES: Remove that service's examples

• If DEVOPS_CONTACTS empty: replace with concise empty-state note, not [TBD: ...]

• Populate service dependencies ONLY from DOCKER_SERVICES

• Populate port mapping ONLY from docker-compose.yml ports section

Phase 3: Self-Validate

For infrastructure.md:

• Check SCOPE tag and metadata markers

• Check required top sections (Quick Navigation , Agent Entry , Maintenance )

• Validate sections: Server Inventory, Port Allocation, Deployed Services

• Check no procedural content leaked (belongs in runbook.md)

• Check docs-quality contract compliance (no forbidden placeholders, no leaked template metadata, valid doc kind/role)

For runbook.md (if created):

• Check SCOPE tag and metadata markers

• Check required top sections (Quick Navigation , Agent Entry , Maintenance )

• Validate sections: Local Development Setup, Deployment, Troubleshooting

• Check env vars documented

• Check docs-quality contract compliance (no forbidden placeholders, no leaked template metadata, valid doc kind/role)

Phase 4: Return Status

{ "created_files": ["docs/project/infrastructure.md", "docs/project/runbook.md"], "skipped_files": [], "quality_inputs": { "doc_paths": ["docs/project/infrastructure.md", "docs/project/runbook.md"], "owners": { "docs/project/infrastructure.md": "ln-115-devops-docs-creator", "docs/project/runbook.md": "ln-115-devops-docs-creator" } }, "validation_status": "passed" }

Critical Notes

Core Rules

• infrastructure.md: Always created, no condition

• runbook.md: Conditional on hasDocker

• Heavy auto-discovery: Most data from docker-compose.yml, .env.example, package.json, SSH config

• Reproducible: Setup steps must be testable and repeatable

• Idempotent: Never overwrite existing files

• Publishable output: No [TBD: ...] , TODO , or leaked template metadata in DevOps docs

NO_CODE_EXAMPLES Rule (MANDATORY)

Both documents describe inventory/procedures, NOT implementations:

• FORBIDDEN: Full Docker configs, CI/CD pipelines (>5 lines), full nginx configs

• ALLOWED: Command examples (1-3 lines), env var tables, step lists, verification commands

• INSTEAD OF CODE: "See docker-compose.yml"

Stack Adaptation Rule (MANDATORY)

• Commands must match project stack (npm vs pip vs go)

• Link to correct cloud provider docs (AWS/Azure/GCP)

• Never mix stack references (no npm commands in Python project)

Format Priority (MANDATORY)

Tables (env vars, ports, services, servers) > Lists (setup steps) > Text

Runtime Summary Artifact

MANDATORY READ: Load shared/references/docs_generation_summary_contract.md

Accept optional summaryArtifactPath .

Summary kind:

• docs-generation

Required payload semantics:

• worker = "ln-115"

• status

• created_files

• skipped_files

• quality_inputs

• validation_status

• warnings

Write the summary to the provided artifact path or return the same envelope in structured output.

Definition of Done

• infrastructure.md created (always)

• runbook.md created if hasDocker

• Infrastructure: server inventory, ports, services documented

• Runbook: setup steps, deployment, troubleshooting documented

• All env vars from .env.example included in runbook

• Actuality verified: all document facts match current code (paths, functions, APIs, configs exist and are accurate)

• Status returned

Reference Files

• Templates: references/templates/infrastructure_template.md , references/templates/runbook_template.md

• Questions: references/questions_devops.md (Q46-Q51 runbook, Q52-Q55 infrastructure)

Version: 2.0.0 Last Updated: 2025-01-12