| name | eaa-label-taxonomy |
| description | GitHub label taxonomy reference for the Architect Agent. Use when designing architecture, identifying components, or recommending labels. Trigger with architecture label requests. Use when this capability is needed. |
| metadata | {"author":"emasoft"} |
EAA Label Taxonomy
Overview
This skill provides the label taxonomy relevant to the Architect Agent (EAA) role. Each role plugin has its own label-taxonomy skill covering the labels that role manages.
Prerequisites
- GitHub CLI (
gh) installed and authenticated
- Active issue number for architecture work
- Understanding of EAA role responsibilities (see AGENT_OPERATIONS.md)
- Completed architecture analysis for the feature
Instructions
- Analyze requirements to identify all affected components
- Review priority labels to understand design constraints
- Check effort labels against design complexity
- Recommend component labels based on module breakdown
- Create sub-issues with appropriate type and component labels
- Validate effort estimates and recommend changes if needed
Checklist
Copy this checklist and track your progress:
Architecture Label Workflow:
Architecture Handoff Checklist:
Output
| Output Type | Format | Example |
|---|
| Component recommendations | Issue comment | Affected components: api, database, auth |
| Label updates | CLI stdout | ✓ Labels updated for #456 |
| Sub-issue creation | Issue URL | Created #457 for API changes |
| Effort validation | Issue comment | Recommend upgrading effort:s → effort:m |
Error Handling
| Error | Cause | Solution |
|---|
label not found | Component label doesn't exist | Create label first via gh label create |
permission denied | No write access to repo | Verify GitHub token scopes |
issue not found | Invalid issue number | Verify issue number with gh issue list |
duplicate label | Label already applied | No action needed, continue |
Labels EAA Manages
Component Labels (component:*)
EAA recommends component labels during architecture design.
| Label | Description | When EAA Recommends It |
|---|
component:api | API endpoints | Feature touches REST/GraphQL APIs |
component:ui | User interface | Feature has UI changes |
component:database | Database/storage | Schema or query changes |
component:auth | Authentication | Auth/authorization changes |
component:infra | Infrastructure | DevOps/deployment changes |
component:core | Core business logic | Central logic changes |
component:tests | Test infrastructure | Test framework changes |
component:docs | Documentation | Doc system changes |
EAA Component Responsibilities:
- Analyze requirements to identify affected components
- Recommend component labels in handoff to EOA
- Update component labels when design changes
Kanban Columns (Canonical 8-Column System)
The full workflow uses these 8 status columns:
| # | Column Code | Display Name | Label | Description |
|---|
| 1 | backlog | Backlog | status:backlog | Entry point for new tasks |
| 2 | todo | Todo | status:todo | Ready to start |
| 3 | in-progress | In Progress | status:in-progress | Active work |
| 4 | ai-review | AI Review | status:ai-review | Integrator agent reviews ALL tasks |
| 5 | human-review | Human Review | status:human-review | User reviews BIG tasks only (via EAMA) |
| 6 | merge-release | Merge/Release | status:merge-release | Ready to merge |
| 7 | done | Done | status:done | Completed |
| 8 | blocked | Blocked | status:blocked | Blocked at any stage |
Task Routing Rules:
- Small tasks: In Progress -> AI Review -> Merge/Release -> Done
- Big tasks: In Progress -> AI Review -> Human Review -> Merge/Release -> Done
- Human Review is requested via EAMA (Assistant Manager asks user to test/review)
- Not all tasks go through Human Review -- only significant changes requiring human judgment
Type Labels EAA Clarifies
EAA may recommend type changes based on architecture analysis:
| Scenario | Type Recommendation |
|---|
| "Add feature" requires refactoring first | type:refactor for prep issue |
| Feature spans multiple systems | type:epic parent + type:feature children |
| Security implications discovered | Add type:security issue |
Labels EAA Reads (Set by Others)
Priority Labels (priority:*)
EAA uses priority to scope architecture:
priority:critical - Minimal viable design, fastest path
priority:high - Solid design, balance speed/quality
priority:normal - Full architecture consideration
priority:low - Can consider future extensibility
Effort Labels (effort:*)
EAA validates effort estimates:
- Does design complexity match effort label?
- Should
effort:m be upgraded to effort:l?
- EAA recommends effort changes to EOA
EAA Label Commands
When Completing Architecture
gh issue edit $ISSUE_NUMBER --add-label "component:api" --add-label "component:database"
gh issue edit $ISSUE_NUMBER --remove-label "effort:s" --add-label "effort:m"
When Creating Sub-Issues
gh issue create \
--title "[$PARENT_ID] API changes for $FEATURE" \
--body "Part of #$PARENT_ISSUE" \
--label "type:feature" \
--label "component:api" \
--label "status:backlog"
When Design Changes Scope
gh issue edit $ISSUE_NUMBER --remove-label "type:feature" --add-label "type:epic"
Architecture-to-Labels Mapping
Module Breakdown → Component Labels
When EAA breaks down a feature:
## Architecture Breakdown
### Feature: User Authentication
**Components Affected:**
- API layer (new endpoints) → `component:api`
- Database (user schema) → `component:database`
- Auth module (new) → `component:auth`
- UI (login form) → `component:ui`
**Recommended Labels:**
- `component:api`
- `component:database`
- `component:auth`
- `component:ui`
Complexity Analysis → Effort Labels
| Architecture Complexity | Effort Recommendation |
|---|
| Single component, clear pattern | effort:s |
| 2-3 components, existing patterns | effort:m |
| Multiple components, new patterns | effort:l |
| System-wide, new architecture | effort:xl |
Examples
Example 1: Adding Component Labels After Architecture Analysis
gh issue edit 123 --add-label "component:api" --add-label "component:database"
Example 2: Validating Effort Estimate
gh issue comment 123 --body "Architecture analysis suggests effort:m (3 components: API, DB, Auth)"
gh issue edit 123 --remove-label "effort:s" --add-label "effort:m"
Example 3: Creating Component-Specific Sub-Issues
gh issue create \
--title "[#123] API endpoints for user authentication" \
--body "Part of #123 - Implements REST API for auth flow" \
--label "type:feature" \
--label "component:api" \
--label "status:backlog" \
--label "effort:s"
Example 4: Architecture Decision Record (ADR)
gh issue create \
--title "[ADR-005] PostgreSQL vs MongoDB for user storage" \
--body "Evaluating database options..." \
--label "type:docs" \
--label "component:database" \
--label "priority:high"
Quick Reference
EAA Label Responsibilities
| Action | Labels Involved |
|---|
| Analyze requirements | Read type:*, priority:* |
| Identify components | Recommend component:* |
| Validate effort | Recommend effort:* changes |
| Create sub-issues | Set type:*, component:*, status:backlog |
| Design change | May update type:* (with EOA approval) |
Labels EAA Never Sets
assign:* - Set by EOA/ECOS
status:* - Set by working agent
review:* - Managed by EIA
priority:* - Set by EAMA/EOA
EAA Handoff Labels
When handing off design to EOA, EAA should ensure:
- All
component:* labels are set
effort:* is validated
- Sub-issues created if
type:epic
ADR Labels Pattern
When creating Architecture Decision Records:
gh issue create \
--title "[ADR-001] Database choice for user storage" \
--body "Architecture decision record..." \
--label "type:docs" \
--label "component:database" \
--label "priority:high"
Resources
- AGENT_OPERATIONS.md - EAA role definition and responsibilities
- eaa-modularization - Module breakdown procedures
Converted and distributed by TomeVault — claim your Tome and manage your conversions.