OpenLogos ships with 18 AI Skills — structured Markdown instruction files (SKILL.md) that guide AI coding tools through each phase of the development lifecycle. Skills are deployed automatically during openlogos init and synchronized via openlogos sync.
Each Skill is a self-contained instruction set stored in skills/<skill-name>/SKILL.md. When an AI coding tool encounters a task matching a Skill’s trigger conditions, it reads and follows the instructions to produce consistent, high-quality outputs.
Skills are organized to follow the OpenLogos three-layer progression model (WHY → WHAT → HOW):
| Skill | Purpose | Trigger |
|---|
project-init | Initialize project structure and configuration | New project setup or openlogos init |
prd-writer | Write scenario-driven requirements with GIVEN/WHEN/THEN criteria | Requirements analysis phase |
| Skill | Purpose | Trigger |
|---|
product-designer | Create interaction specs and prototypes adapted to product type | Requirements exist, design needed |
| Skill | Purpose | Trigger |
|---|
architecture-designer | Design system architecture and select technology stack | Product design complete, Step 0 |
scenario-architect | Expand scenarios into technical sequence diagrams | Architecture complete, Step 1 |
api-designer | Design OpenAPI specs derived from sequence diagrams | Sequence diagrams complete, Step 2 |
db-designer | Derive database DDL from API specifications | API specs complete, Step 2 |
test-writer | Design unit test and scenario test cases | Specs complete, Step 4a (all projects) |
test-orchestrator | Design API orchestration test scenarios | Test cases complete, Step 4b (API projects only) |
code-implementor | Generate business code and test code with spec fidelity | Test design complete, Step 5 |
code-reviewer | Review code against the full specification chain | After code generation, Step 5+ |
| Skill | Purpose | Trigger |
|---|
deployment-designer | Design deployment topology, release commands, rollback strategy, and smoke scope | API/DB complete, Step 3 |
deployment-executor | Execute deployment with human authorization after verify passes | After Gate 3.6 PASS, Step 7 |
| Skill | Purpose | Trigger |
|---|
change-writer | Write change proposals with impact analysis | Any iteration during launched lifecycle |
slice-planner | Plan post-spec-complete [code] slices from merged specs and real test IDs | After merge / no-delta spec-complete, before code implementation |
merge-executor | Merge delta files into main documents | After openlogos merge <slug> |
| Skill | Purpose | Trigger |
|---|
brownfield-adopter | Optional: reverse-engineer a verifiable baseline seed (system map + scenario candidates) from an existing codebase and commit it with openlogos baseline-seed | Only when you explicitly ask for an up-front full-codebase scan; never triggered just because openlogos adopt finished |
| Skill | Purpose | Trigger |
|---|
ui-ux-pro-max | Comprehensive UI/UX design guide (67 styles, 96 palettes, 57 font pairings, 25 chart types) | GUI product design in Phase 2, invoked by product-designer |
After openlogos init, Skills are deployed to platform-specific locations:
| AI Tool | Skill Location |
|---|
| Claude Code | logos/skills/ (native plugin) |
| OpenCode | logos/skills/ (hooks integration) |
| Codex | .agents/plugins/openlogos/skills/ ($openlogos:<skill> native Codex Skills) |
| Cursor | .cursor/skills/<name>/SKILL.md (native Agent Skills) |
| ZCode | .zcode/plugins/openlogos/skills/ (native plugin) |
| Qoder | .qoder/plugins/openlogos/skills/ (native plugin) |
| WorkBuddy | .workbuddy/plugins/openlogos/skills/ (native plugin) |
| Other | logos/skills/ |
Multi-file Skill assets (such as ui-ux-pro-max) always stay under logos/skills/. Run openlogos sync to re-deploy Skills after updates. See init for the full list of files each host receives.
