Future Skill Roadmap

This page turns the current documentation work into a practical roadmap for future Cabloy skills.

Why a roadmap helps

The docs now describe a large portion of the Cabloy backend, frontend, fullstack, and edition-aware workflow surface.

A roadmap helps convert that documented knowledge into a focused set of high-value skills rather than a random collection of prompts.

What already exists

Current root skills include:

• cabloy-workflow
• cabloy-domain-planning
• cabloy-spec-generation
• cabloy-spec-execution
• cabloy-backend-scaffold
• cabloy-frontend-scaffold
• cabloy-contract-loop
• cabloy-resource-field-update
• cabloy-module-removal
• cabloy-master-detail
• cabloy-worktree-environment
• cabloy-zova-source-reading

Their current roles are:

• cabloy-workflow → broad workflow selection, edition detection, CLI-first routing, and verification framing
• cabloy-domain-planning → suite-first naming proposals, providerId confirmation, module-boundary suggestions, and scaffold handoff for new business domains
• cabloy-spec-generation → repository-native suite planning authority, traceability, confirmation, and derived planning views
• cabloy-spec-execution → one approved WBS increment through readiness gates, specialist routing, evidence, and derived status
• cabloy-backend-scaffold → Vona implementation routing and CLI-backed backend scaffolding
• cabloy-frontend-scaffold → Zova implementation routing and CLI-backed frontend scaffolding
• cabloy-contract-loop → backend/frontend contract regeneration, reverse-chain handling, and drift diagnosis
• cabloy-resource-field-update → existing backend resource-field changes with fileVersion and renderer-aware follow-up
• cabloy-module-removal → backend/frontend/fullstack module deletion order, generated-runtime cleanup, and verification
• cabloy-master-detail → coordinated master-detail resource workflow
• cabloy-worktree-environment → confirmed worktree-local environment tuple setup for intentional parallel work
• cabloy-zova-source-reading → Zova-first source reading and Vue comparison workflow

This is now a stronger foundation skill set, but it still leaves several useful workflow families for future specialization.

Recommended next skill families

The backend scaffold, frontend scaffold, contract-loop, resource-field-update, and metadata-refresh workflow families already have implemented specialist coverage. Future work should refine those skills only when a new repository-wide branch or stable verification need emerges.

1. Distributed backend workflow skill

Purpose:

• route tasks into queue / schedule / broadcast / redlock / worker logic
• help choose the right distributed abstraction
• verify mode/flavor and transaction/cache implications

Primary dependencies:

• backend distributed docs
• Redis / queue / schedule / worker / broadcast / redlock pages

What should stay in docs rather than becoming a skill

Not every good doc topic should become a skill.

Keep something in docs only when:

• it is mostly conceptual
• it has no procedural branching worth automating
• it primarily teaches architecture rather than driving action

Examples:

• high-level architecture foundations
• broad conceptual comparisons
• maintainership rationale

Skill design rules for future Cabloy skills

Future skills should generally:

1. detect edition first when relevant
2. classify backend/frontend/fullstack/docs/distributed layer
3. prefer CLI/refactor/generator paths
4. read docs only where the explanation adds value
5. verify the result using the smallest correct command set

Why this roadmap matters for AI workflows

The purpose of the roadmap is not to create many skills.

The purpose is to create a small number of high-leverage skills that directly reuse the knowledge system now being built in:

• repo-docs/
• CLAUDE.md
• repo-docs-internal/
• .claude/skills/
• the Vona and Zova CLIs

That is how Cabloy gets long-term value from the documentation work.