morkit-plugin

Sync task changes into a convention-driven knowledge pack (catalog.json + repos/*.md) and keep a weekly ledger of synced tasks. Manual, batch cadence — run by PM/team lead. Scans gRPC (.proto), REST (Gin), models (GORM), Makefile from source; human-gated via checkbox proposal. No prose auto-edits.

Generate or update full project documentation suite (SRS / API / DB / system-architecture / code-standards / codebase-summary / design-guidelines — user picks via /morkit:init multi-select gate). Orchestrates 7 sub-skills with conflict-minimal updates from OpenSpec changes, brainstorm plans, or codebase scans. Standards: BrSE ITO Japan (SRS), arc42-lite (arch), Conventional Commits (standards), MADR (guidelines). Supports init / update / sync.

Bridge skill — author a valid ProjectModel JSON (per normalized_schema) from greenfield inputs (parsed customer docs + brainstorm report + user-story list + risk register + clarification answers), then validate it so /morkit:init can render docs/srs.md and friends with no hand-authored JSON. The missing brainstorm→init link for the greenfield pipeline.

Generate or update Software Requirements Specification (SRS) following BrSE standards for ITO Japan. Renders the BrSE template-updated structure (13 sections + 2 appendices: Doc Control, Overview, Business Flow with UC detail, FR detail, Business Rules, Roles & Permissions, NFR with IPA-6 categories + Security/PII, Data Items with retention, External Interfaces, Reports, Acceptance/UAT, Traceability, Open Q&A, Constraints/Assumptions/Risks, Screen Index, Glossary). Init mode generates srs.md + per-screen specs from ProjectModel JSON; update mode applies a Delta to existing docs preserving manual edits.

Stateful guide for /morkit:greenfield — walks the BA/BrSE documentation pipeline G0→G7, runs the owning skill per stage, enforces the 4 human gates, and resumes from state.json. Thin glue: holds NO business logic — every stage delegates to an existing skill (brainstorming, generate-user-stories, gap-risk-analysis, clarification-loop, build-project-model) or to /morkit:init for the final SRS + design docs. Turns customer docs into a validated ProjectModel and a full docs/ set with no hand-authored JSON.

Enter explore mode - a thinking partner for exploring ideas, investigating problems, and clarifying requirements. Use when the user wants to think through something before or during a change.

Use when you need to locate files or gather task context across a codebase before implementing, debugging, or reviewing — spawns parallel Explore agents to find relevant files without loading their contents into your own context.

Use when starting any conversation, OR when the user types a `/morkit:<name>` slash-command pattern (e.g. `/morkit:init`, `/morkit:propose`, `/morkit:setup`) - establishes skill discovery rules and the slash-command bridge for platforms without native slash discovery

Turn the greenfield brainstorm report (+ source manifest) into a standalone user-story-list.md — the S2 deliverable that otherwise only exists buried in SRS §3. Confidence-gated co-creation: self-scores each story (source/completeness/interpretation) + a coverage map, runs scoped human Q&A only where unsure (reusing clarification-loop), then a confirm gate — depth adapts to G1 quality. Two output formats via --format brse|agile: brse = function-list rows aligned 1:1 with SRS §3.1 (JP ITO clients); agile = As-a/I-want/So-that with acceptance criteria. Both map cleanly to ProjectModel FunctionalRequirement / UseCase for the bridge.

Help a manual QA tester author test-case specifications into an Excel workbook. Interactive: understand a feature from docs/images/optional live URL, brainstorm coverage, approve a test-scope gate, then write finished test cases (description, pre-condition, procedure, expected output) into the team's Excel template. Writes specs only — never runs tests. Use when a tester says 'write test cases', 'viết test case', 'create test cases for <feature>', or needs a test-case sheet filled.

Generate or update design-guidelines.md (Design Principles + Patterns + ADRs in MADR format). Init renders from ProjectModel and emits per-ADR stubs at docs/adr/NNN-slug.md; update applies a Delta filtered for guidelines scope (DPR/PTN/ADR). Sync mode is intentionally not supported — guidelines are manual.

Archive a completed morkit change. Moves it from active to archive subfolder and updates .meta.json. Use when implementation is complete and merged.

The S4 human-in-the-loop for /morkit:greenfield. AI generates clarifying questions per user-story/gap into clarification-log.md; BA/BrSE answers inline or marks forwarded→<stakeholder>; on re-run AI ingests answers, updates the ProjectModel OpenQuestion/FR detail, and fills <TBD> placeholders. Async + multi-session: all state lives in the markdown file so the loop resumes from the file alone. Closed Q&A renders into SRS §12.

Use when you have a written implementation plan to execute in a separate session with review checkpoints

Read the greenfield user-story list + brainstorm report (+ source manifest) and emit two standalone canonical artifacts — gap-analysis.md and risk-register.md — with fixed templates and numeric risk scoring (H/M/L→3/2/1, Score=Prob×Impact, High≥6 requires mitigation). These are the G3 BA-review gate, then sync into ProjectModel so SRS §13 (Risks) and §12 (Open Q&A) render from the same data.

Generate or update REST API documentation. Init mode renders api-docs.md from ProjectModel; update mode applies a Delta to existing docs; sync is a 2-step propose→apply that scans the codebase and asks the user to pick which discoveries to apply.

Generate or update database design document with Mermaid ERD. Init mode renders database-design.md from ProjectModel; update mode applies a Delta to existing docs; sync is a 2-step propose→apply that scans ORM models in the codebase.

Use when implementation is complete, all tests pass, and you need to decide how to integrate the work - guides completion of development work by presenting structured options for merge, PR, or cleanup

Generate or update code-standards.md (Conventional Commits + auto-extracted lint/format rules). Init renders from ProjectModel; update applies a Delta filtered for standards scope (LNT/NAM/CMT/FMT); sync scans the repo for lint/format/commit configs and proposes Add/Deprecate per rule. Honors existing CONTRIBUTING.md by linking, not duplicating.

Generate or update codebase-summary.md (README-style overview: tech stack, repo layout, packages, entry points, LOC by language). Init renders from ProjectModel; update applies a Delta filtered for summary scope (RPO/TCH/PKG/MOD); sync scans the repo for file tree + manifests + LOC and proposes Add/Deprecate.

Generate or update system architecture document (arc42-lite, 8 sections) with embedded Mermaid component diagram. Init renders system-architecture.md from ProjectModel; update applies a Delta filtered for arch scope (CMP/LAY/INX/QG); sync scans the codebase (services / packages / Dockerfile / k8s / import graph) and proposes Add/Deprecate per Component.

Generate the developer review checklist (human gate) for a morkit change, blocking implementation skills until the human flips "Overall Decision: OK". Triggered by /morkit:review or after /morkit:propose finishes.

Use when the user asks to commit, push, create a PR, or merge branches. Handles conventional commits, secret scanning, and auto-splits commits by type/scope.

Use when executing implementation plans with independent tasks in the current session

Propose a new morkit change with all artifacts generated in one step. Use when the user wants to describe what they want to build and get a complete proposal with design, tasks, and review-checklist ready for implementation. No OpenSpec or external CLI dependency — plugin scaffolds directly.

Use when implementing any feature or bugfix, before writing implementation code

Run a deep multi-agent code review on a PR or git diff. Dispatches 5 specialist subagents (risk, security, pattern, tests, convention) in parallel and synthesizes a Markdown matrix report. Honors project CLAUDE.md as the highest source of truth.

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes

Use when you have a spec or requirements for a multi-step task, before touching code

Use when starting feature work that needs isolation from current workspace or before executing implementation plans - creates isolated git worktrees with smart directory selection and safety verification

Use when completing tasks, implementing major features, or before merging to verify work meets requirements

Use when creating new skills, editing existing skills, or verifying skills work before deployment