Skip to main content

Wire Framework Tutorials

These tutorials are scenario-based walkthroughs of each Wire release type, built around fictional client engagements. They go further than the release-type reference pages: each one shows realistic command output, how agent delegation plays out in practice, which MCP integrations activate at each gate, and the decision-making context that shapes what gets generated. The scenarios are constructed to surface the parts of each release type that are easy to misread from the reference alone — appetite versus scope in a Shape Up discovery, the two-zone safety model in a platform migration, the seed-based prototyping sequence in a Dashboard First release.

The supplementary tutorials at the bottom of the table cover the operational mechanics: installing and upgrading the plugin, joining an active engagement mid-stream, and upgrading an existing release folder when a newer version of Wire is installed. Read the tutorial for your release type before running /wire:new, or keep it open alongside the terminal during delivery. The supplementary tutorials are most useful when something about the environment or the engagement handover is non-standard.

Tutorial index

TutorialRelease TypeScenarioKey Features Shown
Full Platformfull_platformEversholt Brewing Co — Shopify, BrewMan ERP, and HubSpot into BigQuery + LookerAll six phases end-to-end; parallel dbt agent fan-out; Jira hierarchy creation; decisions.md accumulation across agents
dbt Developmentdbt_developmentVantage Financial Reporting — Stripe, Salesforce, and PostgreSQL into SnowflakeTransformation-only scope; cross-system customer identity resolution; 38 schema tests; skipping pipeline and BI phases
Pipeline and dbtpipeline_onlyMeridian Logistics Group — complex multi-source ingestion with a bespoke SFTP connectorConnector configuration and activation; custom Cloud Function pipeline; staging-layer focus before warehouse design is committed
Discovery (Shape Up)discovery_shape_upHallmark Property Partners — real estate investment go/no-go scopingAppetite document; scope story map; risk catalogue; Fathom transcript integration; SOW-ready output in two days
Discovery (SOP)sop_discoveryThornfield Private Healthcare — four-clinic GDPR-sensitive assessmentFormal stakeholder interviews with MoSCoW categorisation; data inventory; capability assessment; Jira Epic auto-created at /wire:new
Dashboard Extensiondashboard_extensionFoxwood Commerce Ltd — marketing dashboard expansion on an existing Looker instanceStarting from semantic_layer-generate; existing LookML pattern-matching; no pipeline or dbt work in scope
Dashboard Firstdashboard_firstClaybrook Media Group — interactive HTML mockup before any data layer is committeddashboard-mock-developer agent; Chart.js interactive prototypes; atomic derivation of viz catalog, dashboard spec, and data model requirements; CSV seed files with referential integrity; mock-to-real refactor sequence
EnablementenablementHargreave Insurance Ltd — platform enablement and technical handoverdelivery-lead agent reading prior build artifacts; two-audience training generation; architecture and field-catalogue documentation
Platform Migrationplatform_migrationGatwick Data Partners — Snowflake to BigQuery migrationAudit zone versus migration zone; five equivalency check types; equivalency-investigate / equivalency-fix loop; cutover gate requiring checks_failing: 0 and written sign-off
Agentic Data Stackagentic_data_stackBoutique analytics consultancy — 47-model sprawl with conflicting metric definitionsPre-agent canonical model audit; knowledge skill authoring; eval suite with CI runner; accuracy regression prevention
DroughtydroughtyBirchfield Capital Management — 240-table Snowflake warehouse, no dbt projectDiscovery/audit mode; DBML entity-relationship diagram from INFORMATION_SCHEMA; AI field descriptions; base LookML view generation; dbt schema test stubs
Custom ReleasecustomSummit Digital Media — content analytics advisory across BigQuery, Looker, and Vertex AIConsultant-defined artifact set; standard generate/validate/review lifecycle applied to bespoke deliverables; Wire infrastructure without a fixed release shape
Installing and UpgradingClaude Code and Gemini CLI installation from scratchThree-step Claude Code install; Gemini CLI extension install; MCP server configuration; verification commands; keeping the plugin current
Joining Mid-ReleaseAldgate Financial Services — consultant handover at Phase 3/wire:start state recovery; decisions.md history; Fathom transcript surfacing; first-session planning on an engagement you did not start
Upgrading Your ReleasePennant Capital Management — dormant release resuming after a six-week pauseWhat /wire:upgrade changes and preserves; spec version delta; post-upgrade verification; re-entrancy with in-progress artifacts
note

The original worked example at Advanced → Worked Example uses a real RA client engagement (Barton Peveril Sixth Form College). These tutorials use fictional scenarios specifically designed to illustrate each release type.