Skip to content

CHANGELOG.md

Latest commit

 

History

History
66 lines (44 loc) · 10 KB

CHANGELOG.md

File metadata and controls

66 lines (44 loc) · 10 KB
Raw

Changelog

All notable changes to bimstack will be recorded here. Format inspired by Keep a Changelog.

[Unreleased]

Added

  • A review skill for any citizen-facing service – service-design-review reviews a form, a calculator or estimator, an entry or information page, or a whole journey against the alpha.gov.bb service patterns and the Barbados field standards. Three lenses (content, flow, standards), output triaged into must fix now / fix later / confirm with the MDA / test with users, and it closes by offering the fixes as a comparison prototype rather than a list of notes. Adds /bimstack:review.

  • Five design-system skills for the GOV.BB Pattern Library – component-spec (reads a component or token layer from Figma and writes an accurate spec: anatomy, states, properties, and the token each part is bound to), token-cross-check (cross-checks the design system across the published Figma library, the live working file, and the front-end code, reporting what is missing, mismatched, or unpublished; reads only, never edits), page-composition (assembles a page or flow in Figma from linked design-system instances – never detached – behind a scope-and-sign-off workflow: scope the flow, gather inspiration as input, scrutinise the service logic, agree the component inventory as a labelled wireframe, fill library gaps first, then compose breakpoint by breakpoint), design-review (reviews a design against the Barbados Digital Service Standards and WCAG 2.1 AA and returns a triaged verdict – Fix now / Fix later / Confirm with MDA – with rigour matched to the surface: measured on a live HTML prototype, inferred on Figma), and component-build (creates a new Pattern Library component bound to real semantic and component tokens and current text styles, never raw values or legacy styles, behind a plan-first gate: it proposes anatomy, variants, and the token each part binds to, and waits for sign-off before it writes). Anchored to Standards 3, 5, and 7

  • Two canonical design references, split so values cannot drift – references/govbb-design-tokens.yaml (the structured, machine-checkable values: the colour, semantic, and component token layers, the type scale, the component inventory, and a scope guard separating what is tokenised from what merely exists in Figma) and references/govbb-design-guide.md (the narrative: how each layer is built, the decisions behind it, and the rules for reading and writing Figma safely). Values live only in the YAML; the skills read the YAML for values, the guide for reasoning, then Figma for current state

  • Cross-references wired in: the content-designer agent's deliverables table now points to bimstack:component-spec, bimstack:token-cross-check, bimstack:page-composition, bimstack:design-review, and bimstack:component-build, and its design-system note defers to component-spec for reading what the system contains

  • Three field-tested skills from the GovTech Barbados research practice – synthesize-research (raw study data → layered findings package: exec summary, briefing docx, GovTech-style findings deck, and appendix, with a themes sign-off checkpoint before anything is written), govtech-research-plans (a complete three-part research plan as one .docx – research plan, discussion guide, note-taking framework – for usability testing, discovery interviews, or concept testing, grounded in the actual prototype being tested), and service-journey-mapping (service blueprints and journey maps as self-contained GovBB-style HTML pages with inline commenting, built from a transcript, a Miro board, or an interview with the user). These complement the existing native skills: research-planning coaches the thinking behind a discussion guide where govtech-research-plans produces the full session document; journey-map/service-blueprint output canonical Markdown where service-journey-mapping produces publishable, commentable HTML

  • Four user research skills – research-coach (router and mentor across the whole research cycle, session craft including the false close, direct feedback on the researcher's craft), research-planning (decision-linked objectives before any discussion guide, digital literacy warm-up block, behavioural questions over opinion questions, the "if this didn't exist" test), transcript-analysis (behaviour over stated preference, workarounds and friction that changed behaviour, cross-referencing prior transcripts as confirms/contradicts/new/extends, coached interpretation), and research-presenting (finding → "so what", shared sense-making with the delivery team, action-first readout structure, memorable plain language)

  • synthetic-research – generate synthetic personas grounded in real population data and run them against a bimstack prototype (persona-lens critique + optional agentic walkthrough), producing a friction report with cross-cutting themes, question-protocol failures, and specific recruitment suggestions for real testing. Sits between /bimstack:build and real user testing; synthetic-findings.md feeds directly to prototype-iteration as Path B structured feedback. Bundled: references/persona-schema.md (persona template + grounding-data guidance) and references/question-protocol.md (the four-question interrogation applied to every form field). Wired into: service-designer agent, AGENTS.md, WORKFLOW.md (Test), PLAYBOOK.md (rapid loop), MANUAL.md, GETTING-STARTED.md, and the README skill tree

  • Cross-references wired in: discovery-kit and prototype-iteration now defer to the research skills; service-designer agent, AGENTS.md, WORKFLOW.md (Listen), PLAYBOOK.md, MANUAL.md, and the README skill tree updated to match

  • Five native service-design skills – journey-map (steps, channels, feelings, pain points, all evidenced or marked [ASSUMPTION]), service-blueprint (frontstage/backstage/systems swimlanes separated by the lines of interaction, visibility, and internal interaction; time and failure-point rows), ecosystem-map (actors, channels, systems, and flows around a citizen need, informal actors included, Standard 7 note per system), experience-map (the citizen's whole goal across services, seams as first-class findings, scope recommendation), and workshop-facilitation (decision-linked working sessions with agenda, capture plan, and facilitation craft). Mapping skills output canonical Markdown plus an optional single-file GovBB-style HTML visual, matching the prototype pattern

Fixed

  • Dead skill references. The service-designer agent, delivery-manager agent, discovery-kit, /bimstack:discover, WORKFLOW.md, and AGENTS.md deferred to service-design:* and design:* skills that ship in separate Anthropic plugins and weren't declared as dependencies – a user installing only bimstack hit dead ends at the core service-design moments. All such references now point to the native bimstack:* skills, with the Anthropic plugins repositioned as optional extensions in MANUAL.md

[0.1.1] – 2026-05-16

Fixed

  • Plugin manifest: removed the agents, commands, and skills enumeration arrays from .claude-plugin/plugin.json. Claude Code auto-discovers these from the agents/, commands/, and skills/ directories – the explicit listings were redundant and may have interfered with registration on some installs.
  • Documentation: every /command reference now uses the /bimstack:command namespaced form that Claude Code actually exposes. The bare /command form does not appear in / autocomplete (plugin commands surface under the /bimstack: prefix).
  • Documentation: agent invocation guidance corrected. Claude Code agents are auto-routed by task description, not @-mentioned. Use /agents to list available agents.

[0.1.0] – 2026-05-16

First public cut. Built end-to-end in a single working session.

Added

  • Five agents – service-designer, content-designer, delivery-manager, developer, cyber-engineer
  • Five skills – service-standard-assessment, plain-language-check, weeknote, discovery-kit, show-the-thing
  • Six slash commands – /bimstack:weeknote, /bimstack:assess, /bimstack:discover, /bimstack:show, /bimstack:plain-language, /bimstack:threat-model
  • Four canonical references – the 13 Barbados Digital Service Standards, the 10 GOV.UK Design Principles, the GDS Way phases (Discovery → Alpha → Beta → Live), and the GovTech Barbados house style
  • Top-level documentation – README, ETHOS, AGENTS, WORKFLOW, CONTRIBUTING
  • Plugin manifest at .claude-plugin/plugin.json
  • MIT licence

Anchored to

Known gaps

These are deliberate omissions for v0.1 – flagged here so the next version knows where to look.

  • No example deliverables in the repo yet. v0.2 should ship at least one worked example per skill.
  • No MCP server registration. Agents call existing skills directly; later versions may bundle bespoke MCPs (e.g. a Trident ID dev sandbox, a design-system component fetcher).
  • No CI yet. v0.2 should lint plain-language usage and check that every agent cites the references it claims to.
  • No hooks. Phase-gate hooks (e.g. "block a Met rating without evidence") are a candidate for v0.3.

Inspirations

Format inspired by Garry Tan's gstack. Worldview inspired by the UK Government Digital Service, particularly Russell Davies and Giles Turnbull on working in the open, and the Cabinet Office Service Manual on agile delivery.