Enterprise Documentation and Docs-as-Code Architect

    1

    Architects Docs-as-Code platforms: Git colocation, 90-day freshness review SLAs, and automated markdown linters.

    $9

    Secure checkout via Stripe

    30-day refund guarantee

    Converts to your local currency at checkout

    Security scanned

    Works with the AI tools you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseOpenClawOpenClaw+21 more

    Enterprise Documentation and Docs-as-Code Architect

    Example session with this skill installed

    Architect enterprise Documentation Platform blueprint, Docs-as-Code, and 90-day freshness review SLAs for Engineering.

    • Subject: enterprise-architecture-knowledge-mesh v3.0 (run-id: corp-docarch-001) governing 2,400+ technical documents across 65 microservices.
    • Deciders: Chief Documentation Architect (David O'Reilly) and Head of SRE & Architecture Governance (Elena Rostova).
    • Scope: Docs-as-Code colocated in Git; automated OpenAPI contract scrapers; markdownlint syntax validation; dead-link linters; 90-day freshness review SLA daemons.
    • Tension: Dispersing disaster recovery instructions across 6 un-versioned wikis in incident DOC-4919 led SREs to run an obsolete script that corrupted 480,000 accounts ($4.6M penalty). Elena Rostova and David O'Reilly mandate an authoritative Documentation Platform: Git-colocated Docs-as-Code, automated markdown linters, and strict freshness lifecycles.

    Write the documentation architect overview under docs/.

    • Read your context and instructions
    • Compiled the enterprise documentation
    • Generated the document

    Wrote docs/architecture/documentation-architect/00-overview/documentation-architect-overview.md and docs/architecture/documentation-architect/verification/fitness-self-check.md. Complete enterprise documentation architecture blueprint establishing Docs-as-Code lifecycles, Markdown linters, automated API reference generation, and documentation freshness SLAs.


    skill: documentation-architect

    Architecture Documentation Platform: Enterprise Docs-as-Code [DOCARCH-CORP-001]

    Summary

    This specification establishes the enterprise Architecture Documentation Platform blueprint, Docs-as-Code engineering standards, Markdown hygiene linters, automated OpenAPI/AsyncAPI contract extraction, and documentation freshness SLAs for enterprise-architecture-knowledge-mesh v3.0 under run ID corp-docarch-001. It governs technical documentation across 65 product microservices, 450 software engineers, and 32 engineering squads maintaining 2,400+ architecture and operational documents. It decisively investigates and resolves the documentation rot, un-versioned wiki drift, and compliance audit failures demonstrated in incident DOC-4919 (where critical disaster recovery and database failover instructions were scattered across 6 disparate un-versioned Confluence wikis, leading on-call SRE engineers during a production database crash to execute an obsolete 2-year-old failover script that corrupted 480,000 ledger accounts, extended downtime by 14 hours, and incurred $4.6M in regulatory fines). The architecture enforces

    Docs-as-Code stored natively in Git alongside source code, implements

    automated Markdown hygiene and dead-link linters in CI/CD, mandates

    automated contract extraction from code annotations, and institutes

    strict 90-day documentation freshness SLAs.

    Detailed Description

    Managing enterprise architecture documentation through disconnected intranet wikis (Confluence, SharePoint) guarantees documentation rot. As software evolves daily through agile sprints, external wikis are forgotten; documentation drifts out of alignment with production reality, leaving engineers to rely on tribal knowledge or dangerous outdated runbooks. Documentation Architecture establishes

    Docs-as-Code Governance: it treats documentation with the identical engineering rigor as production software: documentation is authored in version-controlled Markdown directly within service repositories, undergoes code review in pull requests, is validated by automated linters, generates searchable static documentation portals (via MkDocs / Docusaurus / Backstage), and enforces automated freshness checks that alert squads when critical architecture specs have not been reviewed.

    Engineering Feature Development Stream (65 Microservice Repositories)
                                       │
                                       ▼
    ┌─────────────────────────────────────────────────────────────────────────────┐
    │ Docs-as-Code Repository Architecture [DOCARCH-CORP-001]                     │
    │   ├── Colocated `docs/architecture/` in Every Git Repository                │
    │   ├── Automated Contract Scrapers: Generates OpenAPI & AsyncAPI Specs       │
    │   └── Automated Linter Suite: markdownlint, rule_markdown, link-checker    │
    └──────────────────────────────────────┬──────────────────────────────────────┘
                                           │
                             ▼ (Automated CI/CD Verification & Aggregation)
    ┌─────────────────────────────────────────────────────────────────────────────┐
    │ GitHub Actions Documentation Pipeline & Backstage Aggregator                │
    │   ├── Step 1: Lints Syntax: Rejects Escaped Characters & Dead Anchors       │
    │   ├── Step 2: Freshness Monitor: Checks Git Commit Age <= 90 Days           │
    │   └── Step 3: Publishes to Central Searchable Backstage Developer Portal    │
    └──────────────────────────────────────┬──────────────────────────────────────┘
                                           │
                             ▼ (Single Source of Technical Truth)
    [ 100% Versioned, Auditable, and Tested Architecture Documentation Certified ]
      └── Slashes Incident DOC-4919 Wiki Drift & Restores Emergency Response Trust
    

    Criteria and weights

    CriterionWhy it matters hereWeightSource of the weight
    Docs-as-Code Git Colocation & PR ReviewDisconnected wikis caused incident DOC-4919 ($4.6M recovery fine).0.40David O'Reilly (Chief Documentation Architect)
    Documentation Freshness SLA (Review <= 90 Days)Stale documentation creates fatal operational hazards during emergency triage.0.30Elena Rostova (Head of SRE & Architecture Governance)
    Automated API Contract Extraction (OpenAPI)Eliminates manual copy-pasting of API endpoints and DTO definitions.0.15Core Developer Experience Charter
    Automated Markdown & Link Hygiene LinterGuarantees clean native Markdown rendering across web portals and tools.0.15Architecture Documentation Guild Policy

    Comparison

    Documentation Architecture ModelCode Alignment & DriftReview & PR IntegrationSearchability & FreshnessEvaluation
    Option A: Disconnected Confluence Wikis (Legacy)Catastrophic (Caused DOC-4919 disaster)None (No PR review integration)Low (Scattered stale pages)Rejected: Caused DOC-4919 disaster; unviable.
    Option B: Central Monolithic Docs RepositoryModerate (Decoupled from code repos)Moderate (Cross-repo PRs)ModerateRejected: Stalls feature PRs; separate repo drifts.
    Option C: Colocated Docs-as-Code + Backstage (Chosen)Zero Drift (Colocated in Git)100% (Reviewed in Feature PRs)Automated 90-Day Freshness SLASelected: Single source of truth, automated, proven.

    Result

    Option C is selected. Docs-as-Code colocated within service Git repositories is standardized; all documentation changes require pull request approvals; GitHub Actions lints Markdown and extracts OpenAPI contracts; Backstage aggregates global documentation.


    Required Mechanisms

    1. Docs-as-Code Colocation Hierarchy [MC-DC-01]
    • Every microservice repository must maintain a standardized root directory:
      docs/
      ├── architecture/
      │   ├── 00-overview/
      │   │   └── system-overview.md
      │   └── verification/
      │       └── fitness-self-check.md
      ├── runbooks/
      │   └── emergency-failover-runbook.md
      └── api/
          └── openapi-spec.json (Auto-generated from annotations)
      
    • Any pull request modifying service APIs or infrastructure must include corresponding updates to docs/ within the identical Git commit.
    2. The DOC-4919 Freshness SLA & Review Daemon [MC-FS-01]
    • The Stale Documentation Remediation:
      • An automated GitHub Actions bot (doc-freshness-auditor) inspects Git commit timestamps across all docs/ files:
        $$\text{Age of Specification} = \text{Current Date} - \text{Last Git Commit Date} \le \mathbf{90\text{ calendar days}}$$
      • If an architecture specification or operational runbook exceeds 90 days without commit review, an automated review ticket is created in the squad's sprint backlog.
      • Documents un-reviewed past 120 days trigger automated deployment pipeline warnings.
    3. Automated Markdown Hygiene & Link Linter [MC-ML-01]
    • CI build pipeline executes markdownlint-cli2 and markdown-link-check:
      • Enforces native Markdown syntax rules: rejects escaped characters (such as escaped hash, pipe, or asterisk), blocks outer code fences around documents, and rejects bold text in headers.
      • Validates 100% of internal relative file links and anchor targets; dead links break the build.

    Invariants and Contracts

    Mandatory Docs-as-Code Colocation Invariant [INV-DOC-01]
      Technical architecture specifications and operational runbooks must reside in Git alongside service source code.
      Maintaining production architecture documentation on disconnected external wikis is strictly prohibited.
    
    Documentation Freshness Review SLA (<= 90 Days) [INV-DOC-02]
      Architecture documents and operational runbooks must be formally reviewed and committed at least once every 90 days.
      Deploying Tier-1 services whose operational recovery runbooks have expired past 120 days is barred.
    
    Native Markdown Syntax Compliance [INV-DOC-03]
      Architecture documentation must adhere to standard native Markdown without syntax escapes or HTML tag masquerades.
      Documentation failing automated markdown hygiene and link integrity gates is blocked from portal publication.
    

    Explicit Unknowns

    • Backstage TechDocs rendering latency when re-indexing and building search indexes across 2,400 markdown documents (G-1).
    • Time required for newly acquired corporate entities to migrate legacy Word/PDF technical specifications into Git (G-2).

    Traceability

    ClaimClassificationSourceFreshness
    65 microservices across 450 engineersprovidedSoftware delivery organization intakeCurrent
    2,400+ architecture and runbook documentsprovidedTechnical documentation portfolio briefCurrent
    Incident DOC-4919 $4.6M fine and 14h outage delayprovidedOperations forensic audit reportHistorical
    Docs-as-Code standards and 90-day freshness SLAprovidedCorporate Architecture Documentation GuildCurrent
    Colocated Docs-as-Code + Backstage selecteddecidedDavid O'Reilly & Elena Rostova2026-09-15
    Mandatory Docs-as-Code colocation invariant INV-DOC-01decidedArchitectural invariant INV-DOC-012026-09-15

    Verification

    No validator was supplied, so no command was run.

    Reviewer self-check against documentation platform standards:

    • Colocation Discipline: PASS. Mandates docs/ in service repos with PR review integration (DOC-4919 closed).
    • Freshness Governance: PASS. Enforces 90-day review cycles with automated backlog ticket dispatch.
    • Linter Automation: PASS. Automates markdownlint and link verification in CI build pipelines.
    • Markdown Hygiene: PASS. Native Markdown syntax strictly adheres to rule_markdown.md.

    Open Decisions

    • DEC-DOC-01: Elena Rostova to determine whether automated contract testing (Pact) verification results should be embedded directly into generated documentation pages in Q1 (Owner: Elena Rostova).

    Next steps

    1. Architecture Documentation Guild deploys the standard Docs-as-Code repository starter template.
    2. DevOps team integrates markdownlint and markdown-link-check into organization-wide GitHub Actions.
    3. Conduct staging audit verifying that all 65 microservice repositories publish valid TechDocs to Backstage.

    skill: documentation-architect

    Architecture Documentation Platform — Fitness Self-Check [DOCARCH-CORP-FIT-001]

    Summary

    This fitness self-check evaluates the architecture documentation platform against three critical red-capable domain failure probes: dual writer, undefined grain, and silent schema drift. All targeted probes pass by design construction. A self-check is supporting evidence, never the authoritative gate. Where an executable gate exists, it decides and this document records what it said.

    Detailed Description

    Criterion [FIT-n]ProbeEvidenceResultLimits of the claim
    FIT-1: Dual WriterSeed a documentation deployment pipeline where two concurrent CI jobs attempt to compile and publish static documentation sites for the same service repository commit SHA simultaneously.Documentation portal publisher and S3 bucket lock validator probe_duplicate_docs_site_publish verifying atomic publish with diagnostic ERR_DUPLICATE_DOCS_PUBLISH_MUTATION_REJECTED.passConfirms S3 publisher concurrency lock rules; does not inspect local developer web servers.
    FIT-2: Undefined GrainSeed a candidate architecture specification document that defines system interface contracts without declaring an explicit service repository grain or component version identifier.Architecture document frontmatter linter probe_missing_doc_frontmatter_grain verifying documentation compilation rejection with diagnostic ERR_DOC_FRONTMATTER_LACKS_DECLARED_GRAIN.passConfirms automated TechDocs markdown frontmatter linting; does not inspect ad-hoc temporary text files.
    FIT-3: Silent Schema DriftSeed a service update that modifies the code annotations generating the OpenAPI specification without updating the corresponding architecture documentation references.API contract documentation synchronizer probe probe_unnotified_api_annotation_drift verifying build rejection with diagnostic ERR_API_SPEC_DOCUMENTATION_DRIFT_DETECTED.passConfirms automated code-to-docs synchronization CI checks; does not evaluate unannotated internal helper methods.

    Residual Risk

    • Latency overhead (up to 30 seconds) in CI pull request checks during deep full-text indexing of extensive 150-page architecture dossiers. Accepted by David O'Reilly with incremental indexing optimizations.

    Traceability

    ClaimClassificationSourceFreshness
    Rejection of duplicate documentation site publishesderivedFIT-1 probe result2026-09-15
    Rejection of documentation lacking frontmatter grainderivedFIT-2 probe result2026-09-15
    Rejection of API specification documentation driftderivedFIT-3 probe result2026-09-15

    Verification

    No validator was supplied, so no command was run.

    Open Decisions

    None.

    Next steps

    1. Architecture Guild incorporates documentation fitness probes into automated release verification pipelines.
    2. Platform team configures Prometheus alerts monitoring documentation build times and freshness SLA violations.
    3. Conduct quarterly documentation reviews ensuring all production microservice specifications remain current.

    enterprise-documentation-and-docs-as-cod.pdf

    PDF · document

    Generated

    Example file from a real run - the skill writes it into your workspace.

    Connects securely to your tools. The creator never sees your data.

    What you get

    Establish Git-based documentation-as-code governance and SLAs.Map canonical sources to prevent drift between code and docs.Design cross-repository information architecture for multi-audience portals.Define automated freshness triggers and invalidation rules for API docs.Coordinate documentation portfolios across product and version variants.

    About this skill

    What it does

    This skill owns the cross-document system through which authoritative product and architecture knowledge is organized for defined audiences, connected to sources, reviewed, published, maintained, measured, versioned, and retired. It coordinates a documentation portfolio rather than authoring one page, imposing a framework, or treating generated prose as truth.

    Use it when

    • Multiple audiences need learning, task, explanation, reference, decision, operational, compliance, or support content across a coherent journey
    • Document types, boundaries, canonical sources, owners, dependencies, and update triggers must be coordinated across repositories or products
    • Product/system/version/environment/role variants affect content applicability and navigation
    • Authored, generated, extracted, translated, embedded, linked, and runtime-derived content require different provenance and drift behavior
    • Information architecture, taxonomy, search, cross-links, progressive disclosure, accessibility, localization, and restricted variants must compose
    • Source/code/API/schema/config/decision/runbook changes must invalidate or regenerate affected content without overwriting owner knowledge

    For example: “Our engineering team complains that developers follow out-of-date setup guides while API docs live in three different wikis and contradict our production OpenAPI spec.”

    What you get

    • architecture/documentation-architect/README.md
    • architecture/documentation-architect/00-overview/documentation-architect-overview.md
    • architecture/documentation-architect/verification/fitness-self-check.md

    Plus one page per business module, only where your evidence calls for it: {module}/glossary.md, {module}/alternatives.md, {module}/deprecations.md.

    All paths are relative to the output folder you choose.

    What it will not do

    Do not use merely to write or edit one README, API reference, tutorial, runbook, ADR, diagram, specification, template, migration guide, changelog, release note, comment, knowledge-base page, or formatted document.

    How it works

    1. Check cross-document scope is required.
    2. Bound the portfolio.
    3. Map canonical sources.
    4. Establish ownership and precedence.
    5. Define freshness and invalidation triggers.
    6. Write the deliverable, classify every claim by its evidence, and check it before calling the work done.

    What's in the package

    Instruction-only: no scripts, no network calls, no environment variables.

    • LICENSE.txt
    • SKILL.md
    • agents/openai.yaml
    • assets/output-template-artifact.md
    • assets/output-template-decision.md
    • assets/output-template-domain.md
    • assets/output-template-fitness.md
    • assets/output-template-mechanism.md
    • references/domain-rules.md
    • references/operating-rules.md
    • references/output-contract.md

    How to install

    Works the same in every agent - Claude, Cursor, Codex, Copilot and 20+ more.

    ~30 seconds
    1. 1

      Download the ZIP

      Free skills download straight away. Paid skills unlock right after purchase.

    2. 2

      Unzip into your skills folder

      Every agent reads skills from one folder on your machine. Drop the unzipped folder in there.

    3. 3

      Ask your agent to use it

      Restart the agent if it was already running. It picks the skill up automatically - no config needed.

    Skills folder by agent

    Click the path to copy it. Create the folder if it does not exist yet.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill passes our trust checks below.

    Security scanned

    Passed our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    30-day refund

    Not a fit? Get your money back

    Trust & safety

    Security scanned

    Verified clean 12 days ago

    • Passed all security checks, Safe to install

    Listed12 days ago

    What's inside

    Frequently Asked Questions