
{{ 𝛀𝛀𝛀 }} Review a pull request and post a comment
{{ 𝛀𝛀𝛀 }} Decompose a Linear issue into a critical path of granular issues
{{ ƔƔƔ }} Generate a README for a specific directory
{{ ƔƔƔ }} Update a PR description to account for commits made since it was last written
{{ 𝚫𝚫𝚫 }} Grant a permission globally (all projects) via settings.local.jsonc
{{ ƔƔƔ }} Create a draft pull request to main
{{ 𝛀𝛀𝛀 }} Investigate a codebase in detail and write findings to a document, for a named concept, subsystem, or question
Git workflow: branch management, commit conventions, PR patterns, conflict resolution.
{{ ƔƔƔ }} Summarise what the user can now see or do that they couldn't before this unit of work
{{ 𝚫𝚫𝚫 }} Achieve what I say with a minimalist approach
{{ 𝛀𝛀𝛀 }} Review a pull request
{{ ƔƔƔ }} Generate an HTML dashboard showing the current state of a project roadmap.
{{ ƔƔƔ }} Create a draft pull request to main
{{ ƔƔƔ }} Create a pull request to main
{{ 𝛀𝛀𝛀 }} Create a pull request to main
{{ ƔƔƔ }} Create a pull request to main
{{ 𝛀𝛀𝛀 }} Review a pull request
{{ 𝛀𝛀𝛀 }} Investigate a codebase in detail
{{ ƔƔƔ }} Create a pull request to main with screenshots
{{ ƔƔƔ }} Create a draft pull request to main
{{ ƔƔƔ }} Create a pull request to main
Polyglot persistence: when to use relational, graph, or document databases; integration patterns.
{{ ƔƔƔ }} Create an Architecture Decision Record (ADR) for a significant technical decision
{{ ƔƔƔ }} Update an existing documentation file to reflect recent code changes
{{ ƔƔƔ }} Summarise what the user can now see or do that they couldn't before this unit of work
OpenTUI terminal UI library reference. Use when working with @opentui/core, terminal UIs, renderables, Yoga layouts, or Zig-native rendering.
Anti-scope-creep. Forces the question: what is the smallest thing that delivers value?
Model-first design: map entities, relationships, and boundaries before writing code.
Neo4j and Cypher: graph schema design, query patterns, performance optimisation, PostgreSQL integration.
{{ 𝚫𝚫𝚫 }} Grant a permission rule globally or for the current project
{{ 𝛀𝛀𝛀 }} Merge from a branch (default main)
{{ 𝛀𝛀𝛀 }} Rebase from a branch (default main)
{{ 𝛀𝛀𝛀 }} Squash and rebase onto a branch (default main)
{{ 𝚫𝚫𝚫 }} Grant a permission for the current project via .claude/settings.local.json
{{ 𝛀𝛀𝛀 }} Merge from main
{{ 𝛀𝛀𝛀 }} Rebase from main
{{ 𝛀𝛀𝛀 }} Assess branch readiness for PR submission
{{ 𝛀𝛀𝛀 }} Review a pull request and post a comment
{{ ƔƔƔ }} Create or update a README — project root or any directory
{{ 𝛀𝛀𝛀 }} Add a task to a rich-format project roadmap with correct ID, dependency wiring, and graph integrity — ID assignment, status computation, dependency edges in both directions, and no unconnected islands.
{{ 𝛀𝛀𝛀 }} Create a project roadmap in the rich phase-array format — roadmaps.json as source of truth plus a PHASE task list and prose overview
{{ 𝚫𝚫𝚫 }} Integrate a target branch into the current one by merge, rebase or squash
{{ 𝛀𝛀𝛀 }} Create a new Linear issue with automatic dependency detection
{{ 𝛀𝛀𝛀 }} Squash and rebase
{{ 𝚫𝚫𝚫 }} Generate a commit message. If nothing staged, stage all changes.
{{ 𝚫𝚫𝚫 }} Grant a permission globally (all projects) via settings.local.jsonc
{{ 𝛀𝛀𝛀 }} Merge from branch
{{ ƔƔƔ }} Check the current branch name against convention (type/short-description) and rename it if it drifted — e.g. work started on a misnamed or default branch.
{{ 𝛀𝛀𝛀 }} Run a structured interview to discover new features and produce a batch roadmap proposal. Produces a structured proposal for review — nothing is written to the roadmap until the user approves.
{{ ƔƔƔ }} Create a pull request to main with screenshots
{{ 𝛀𝛀𝛀 }} Initialize a new project from the Kamino template
{{ ƔƔƔ }} Create a pull request to main
{{ ƔƔƔ }} Create a draft pull request to main
{{ ƔƔƔ }} Update a PR description to account for commits made since it was last written
{{ ƔƔƔ }} Create a draft pull request to main
{{ 𝛀𝛀𝛀 }} Review a pull request
{{ 𝚫𝚫𝚫 }} Create a status report that knows what you've done since the last one
{{ 𝚫𝚫𝚫 }} Split uncommitted changes into granular commits.
{{ 𝛀𝛀𝛀 }} Create a pull request to main
{{ 𝚫𝚫𝚫 }} Generate the HTML roadmap dashboard deterministically via roadmap.py render.
{{ ƔƔƔ }} Plan a non-trivial feature by interviewing to resolve unknowns, then studding every function as a runnable walking-skeleton (fake data, real wiring) so names, placement, and contracts can be reviewed before any real logic is written. Stops for review; fill is a separate pass.
{{ 𝛀𝛀𝛀 }} Audit a topic and render an actionable, status-grouped HTML findings artefact.
{{ 𝛀𝛀𝛀 }} Render a visual HTML introduction to this codebase for a newly-joined developer
{{ 𝛀𝛀𝛀 }} Review a pull request
{{ 𝛀𝛀𝛀 }} Probe the project for architectural weaknesses, technical debt, and risk — a deliberately critical read, not a balanced overview
{{ 𝛀𝛀𝛀 }} Convert an exported Claude artefact (HTML or JSX) into a working Svelte 5 / SvelteKit 2 project
{{ ƔƔƔ }} Create a pull request to main — wordy or shiny (with screenshots), ready-for-review or draft
Use when about to build a non-trivial feature and you want the shape reviewable before the logic. Studs every function in its real file as a runnable walking-skeleton (fake data, real wiring) so names, placement, and contracts are visible before any real logic is written. Lay the skeleton, then checkpoint for review.
Systematic debugging methodology — runtime errors, test failures, logic bugs, performance issues, production incidents. Five-step framework, root-cause analysis, browser/Node/Svelte tooling, and common bug patterns.
Type-safe API design: Zod validation, Result types, SvelteKit endpoints, middleware patterns.
{{ 𝛀𝛀𝛀 }} Investigate this repo's dependencies in detail
Pragmatic testing with Vitest: risk-based strategy, Svelte component testing, test-after development.
{{ ƔƔƔ }} Update a PR description to account for commits made since it was last written
Svelte 5 and SvelteKit: runes reactivity, component composition, routing, data loading, form handling.
Passive ethical review: manipulation, accessibility, privacy, sustainability in user-facing features.
{{ 𝛀𝛀𝛀 }} Convert an old simple-style roadmap (single Markdown, four statuses, <a name> anchors, roadmaps.json pointer registry) into the rich phase-array format (roadmaps.json source of truth + PHASE task list + prose overview).
Writing style guide for Jason Warren. Applies whenever writing or editing substantive prose for Jason — drafting sections, reports, documentation, or rewriting passages.
Empathy-driven design: think through how real people experience the software; surface and challenge assumptions.
Frontend styling: layout debugging, style consistency, CSS best practices for Svelte/SvelteKit.
{{ 𝚫𝚫𝚫 }} Achieve what I say with a minimalist approach
{{ 𝛀𝛀𝛀 }} Review a pull request and post it as a GitHub review
{{ 𝚫𝚫𝚫 }} Rebuild roadmap-system.zip, the distributable snapshot of the roadmap tooling (scripts, HTML template, conventions reference, and every roadmap-touching skill, including this one).
{{ 𝛀𝛀𝛀 }} Recompute and synchronise roadmap task statuses across roadmaps.json and its projections, with optional codebase reconciliation
--- name: "Suggest: Task" description: "{{ 𝚫𝚫𝚫 }} Suggest the next logical task — grounded in the roadmap's pre-vetted ready-set when one exists, codebase analysis otherwise" when_to_use: "When you don't know what to work on next and want a grounded recommendation rather than picking arbitrarily." model: haiku effort: low disable-model-invocation: true allowed-tools: ["Read", "Glob", "Grep", "Bash(python3:*)", "Bash(npm:*)", "Bash(bun:*)", "Bash(pnpm:*)", "Bash(deno:*)"] argument-hint: [named
{{ ƔƔƔ }} Create a new Linear issue with optional relations
{{ 𝛀𝛀𝛀 }} Analyse and assign dependency relationships for Linear tasks
{{ 𝛀𝛀𝛀 }} Tackle a task from Linear
{{ 𝛀𝛀𝛀 }} Find Linear issues addressed by this branch and inject into PR body
{{ ƔƔƔ }} Rename branch if needed
{{ 𝛀𝛀𝛀 }} Provide a detailed high-level overview of this codebase
{{ 𝛀𝛀𝛀 }} Merge from main
{{ 𝛀𝛀𝛀 }} Run a structured interview to discover new features and produce a batch roadmap proposal. Use this skill when the user wants to explore what to build next, brainstorm features, expand the roadmap, plan a new phase, or says things like 'what should we add', 'help me think through features', 'let's plan the next milestone', or 'interview me about what to build'. Produces a structured proposal for review — nothing is written to the roadmap until the user approves.
{{ 𝛀𝛀𝛀 }} Create a project roadmap document in structured milestone format
{{ ƔƔƔ }} Suggest the next logical task based on codebase analysis
{{ 𝚫𝚫𝚫 }} Grant a permission for the current project via .claude/settings.local.json
{{ 𝚫𝚫𝚫 }} Suggest the next logical task based on codebase analysis
{{ 𝛀𝛀𝛀 }} Suggest the next logical task based on codebase analysis
{{ 𝛀𝛀𝛀 }} Assess branch readiness for PR submission
{{ 𝚫𝚫𝚫 }} Tackle a task from Linear
{{ 𝚫𝚫𝚫 }} Split uncommitted changes into granular commits.
{{ 𝛀𝛀𝛀 }} Squash and rebase
{{ 𝚫𝚫𝚫 }} Generate a commit message. If nothing staged, stage all changes.
{{ 𝛀𝛀𝛀 }} Generate a comprehensive README.md from project analysis
{{ ƔƔƔ }} Update a README to reflect recent changes
{{ ƔƔƔ }} Tackle a task from Linear
{{ ƔƔƔ }} Generate an HTML dashboard showing the current state of a project roadmap.
{{ 𝛀𝛀𝛀 }} Decompose a Linear project into a critical path of granular issues
{{ 𝛀𝛀𝛀 }} Rebase from main
{{ 𝛀𝛀𝛀 }} Merge from branch
{{ 𝛀𝛀𝛀 }} Initialize a new project from the Kamino template
{{ 𝛀𝛀𝛀 }} Add a task to a project roadmap with correct ID, dependency wiring, and graph integrity. Use this skill whenever the user wants to add a task, feature, or work item to a roadmap — even if they just say 'add this to the roadmap', 'put this in the plan', or 'track this as a task'. Handles task ID assignment, section placement, dependency edges in both directions, and ensures no task is left as an unconnected island.
{{ 𝛀𝛀𝛀 }} Review a pull request and post a comment
{{ ƔƔƔ }} Create a draft pull request to main
{{ 𝛀𝛀𝛀 }} Update project roadmap task organization
--- name: api-designer description: Type-safe API design: Zod validation, Result types, SvelteKit endpoints, middleware patterns. user-invocable: false effort: medium paths: - "**/routes/**" - "**/api/**" allowed-tools: - Read - Glob - Grep --- # TypeScript API Design Comprehensive guide to designing type-safe APIs with TypeScript. Covers type-safe contracts, validation with Zod, Result types for error handling, SvelteKit endpoints, middleware patterns, and API versioning. --- ## W
--- name: cypher-linguist description: Neo4j and Cypher: graph schema design, query patterns, performance optimisation, PostgreSQL integration. user-invocable: false effort: medium paths: - "**/*.cypher" - "**/neo4j/**" allowed-tools: - Read - Glob - Grep --- # Neo4j/Cypher Mastery Comprehensive guide to Neo4j graph database and Cypher query language. Covers fundamental concepts, common patterns, performance optimization, schema design, and integration with PostgreSQL/Supabase. ---
--- name: data-ontologist description: Polyglot persistence: when to use relational, graph, or document databases; integration patterns. user-invocable: false effort: high paths: - "**/schema*" - "**/migrations/**" allowed-tools: - Read - Glob - Grep --- # Polyglot Persistence Architecture Architectural guidance for using multiple database technologies together, with emphasis on PostgreSQL/Supabase (relational), Neo4j (graph), and MongoDB (document). Demonstrates when to use each dat
Systematic debugging methodology — runtime errors, test failures, logic bugs, performance issues, production incidents. Five-step framework, root-cause analysis, browser/Node/Svelte tooling, and common bug patterns.
{{ ƔƔƔ }} Generate a README for a specific directory
{{ 𝚫𝚫𝚫 }} Generate a work record summarizing today's development session
{{ ƔƔƔ }} Update a README to reflect recent changes
{{ 𝛀𝛀𝛀 }} Update project roadmap task organization
--- name: domain-modeller description: Model-first design: map entities, relationships, and boundaries before writing code. user-invocable: false effort: high allowed-tools: - Read - Glob - Grep --- # Domain Modeller Model-first design skill. Maps entities, relationships, and boundaries before code. Prioritises understanding the domain over jumping to implementation. Outputs conceptual models as Mermaid diagrams or Cypher-compatible graph structures. --- ## When This Skill Applies Use
--- name: frontend-styler description: Frontend styling: layout debugging, style consistency, CSS best practices for Svelte/SvelteKit. user-invocable: false effort: medium paths: - "**/*.svelte" - "**/*.css" allowed-tools: - Read - Glob - Grep --- # Frontend Styling Comprehensive guidance for debugging layout issues, ensuring style consistency, and applying best practices in frontend development, with emphasis on Svelte/SvelteKit projects. --- ## When This Skill Applies Use this s
--- name: git-manager description: Git workflow: branch management, commit conventions, PR patterns, conflict resolution. user-invocable: false effort: low allowed-tools: - Read - Bash --- # Git Workflow Patterns Comprehensive git workflow guidance covering branch management, commit conventions, pull request best practices, conflict resolution, and LazyGit integration. Emphasizes clean history, collaboration patterns, and the user's established branch naming conventions. --- ## When This
{{ 𝛀𝛀𝛀 }} Create a new Linear issue with automatic dependency detection
{{ ƔƔƔ }} Create a new Linear issue with optional relations
{{ 𝛀𝛀𝛀 }} Decompose a Linear project into a critical path of granular issues
{{ 𝛀𝛀𝛀 }} Find Linear issues addressed by this branch and inject into PR body
{{ 𝛀𝛀𝛀 }} Merge from main
{{ 𝛀𝛀𝛀 }} Rebase from main
{{ 𝛀𝛀𝛀 }} Squash and rebase
{{ 𝛀𝛀𝛀 }} Assess branch readiness for PR submission
{{ 𝛀𝛀𝛀 }} Add a task to a project roadmap with correct ID, dependency wiring, and graph integrity. Use this skill whenever the user wants to add a task, feature, or work item to a roadmap — even if they just say 'add this to the roadmap', 'put this in the plan', or 'track this as a task'. Handles task ID assignment, section placement, dependency edges in both directions, and ensures no task is left as an unconnected island.
{{ 𝛀𝛀𝛀 }} Suggest the next logical task based on codebase analysis
--- name: testing-obsessive description: Pragmatic testing with Vitest: risk-based strategy, Svelte component testing, test-after development. user-invocable: false effort: medium paths: - "**/*.test.*" - "**/*.spec.*" allowed-tools: - Read - Glob - Grep - Bash --- # Testing Foundations Comprehensive testing guidance for JavaScript/TypeScript applications, with emphasis on Vitest, Svelte component testing, and pragmatic test-after development. Addresses testing as a professional sk
{{ ƔƔƔ }} Summarise what the user can now see or do that they couldn't before this unit of work
{{ 𝛀𝛀𝛀 }} Review code changes on the current branch against its open PR
{{ 𝛀𝛀𝛀 }} Create a project roadmap document in structured milestone format
--- name: ethics-reviewer description: Passive ethical review: manipulation, accessibility, privacy, sustainability in user-facing features. user-invocable: false effort: low allowed-tools: - Read --- # Ethics Reviewer Passive ethical review skill. Activates when user-facing features touch the four ethical concerns: manipulation, accessibility, privacy, and sustainability. Checks designs and implementations against principled constraints without requiring explicit invocation. --- ## When T
{{ 𝚫𝚫𝚫 }} Tackle a task from Linear
{{ 𝛀𝛀𝛀 }} Tackle a task from Linear
OpenTUI terminal UI library reference. Use when working with @opentui/core, terminal UIs, renderables, Yoga layouts, or Zig-native rendering.
{{ 𝛀𝛀𝛀 }} Investigate this repo's dependencies in detail
{{ 𝛀𝛀𝛀 }} Run a structured interview to discover new features and produce a batch roadmap proposal. Use this skill when the user wants to explore what to build next, brainstorm features, expand the roadmap, plan a new phase, or says things like 'what should we add', 'help me think through features', 'let's plan the next milestone', or 'interview me about what to build'. Produces a structured proposal for review — nothing is written to the roadmap until the user approves.
--- name: svelte-ninja description: Svelte 5 and SvelteKit: runes reactivity, component composition, routing, data loading, form handling. user-invocable: false effort: medium paths: - "**/*.svelte" - "**/+page.*" - "**/+layout.*" allowed-tools: - Read - Glob - Grep --- # Svelte/SvelteKit Patterns Comprehensive guide to Svelte 5 and SvelteKit development patterns. Emphasizes runes-based reactivity ($state, $derived, $effect, $props), component composition, SvelteKit routing, data l
{{ 𝛀𝛀𝛀 }} Map out project status, direction, and next steps
Writing style guide for Jason Warren. Use this skill whenever writing prose, reports, documentation, or any substantive text for Jason — including drafting sections, editing existing content, or rewriting passages. Also use when Jason asks you to review or improve writing. Trigger on any request involving writing, drafting, editing, or composing text that isn't purely code. This includes github Pull Requests & Linear tasks
{{ 𝚫𝚫𝚫 }} Achieve what I say with a minimalist approach
{{ ƔƔƔ }} Visualise a project roadmap as an interactive HTML dashboard
{{ ƔƔƔ }} Suggest the next logical task based on codebase analysis
{{ 𝚫𝚫𝚫 }} Grant a permission globally (all projects) via settings.local.jsonc
{{ 𝚫𝚫𝚫 }} Grant a permission for the current project via .claude/settings.local.json
{{ ƔƔƔ }} Create an Architecture Decision Record (ADR) for a significant technical decision
{{ 𝛀𝛀𝛀 }} Generate a comprehensive README.md from project analysis
{{ 𝚫𝚫𝚫 }} Create a status report that knows what you've done since the last one
{{ ƔƔƔ }} Update existing documentation to reflect recent code changes
{{ 𝚫𝚫𝚫 }} Split uncommitted changes into granular commits.
{{ 𝚫𝚫𝚫 }} Generate a commit message. If nothing staged, stage all changes.
{{ ƔƔƔ }} Rename branch if needed
{{ 𝛀𝛀𝛀 }} Analyse and assign dependency relationships for Linear tasks
{{ 𝛀𝛀𝛀 }} Decompose a Linear issue into a critical path of granular issues
{{ ƔƔƔ }} Tackle a task from Linear
{{ 𝛀𝛀𝛀 }} Merge from branch
{{ 𝛀𝛀𝛀 }} Probe the project for weaknesses
{{ 𝛀𝛀𝛀 }} Provide a detailed high-level overview of this codebase
--- name: scope-coach description: Anti-scope-creep. Forces the question: what is the smallest thing that delivers value? user-invocable: false effort: low allowed-tools: - Read --- # Scope Coach Anti-scope-creep skill. Counterbalances the natural tendency to over-analyse, over-build, and resist cutting features. Forces the question: "What's the smallest thing that delivers value?" --- ## When This Skill Applies Use this skill when: - A feature design is growing in complexity - Multiple a
{{ 𝚫𝚫𝚫 }} Suggest the next logical task based on codebase analysis
--- name: user-empathy-lens description: Empathy-driven design: think through how real people experience the software; surface and challenge assumptions. user-invocable: false effort: low allowed-tools: - Read --- # User Empathy Lens Empathy-driven design skill. Helps think through how real people will experience the software. Uses inference and scenario-building rather than formal user research — surface assumptions, challenge them, and design for actual human behaviour. --- ## When This
{{ 𝚫𝚫𝚫 }} Check all version number props and update them
{{ 𝛀𝛀𝛀 }} Initialize a new project from the Kamino template